Frederick Haselton+2007

Personality and Social Psychology Bulletin
http://psp.sagepub.com
Why Is Muscularity Sexy? Tests of the Fitness Indicator Hypothesis

David A. Frederick and Martie G. Haselton
Pers Soc Psychol Bull 2007; 33; 1167 originally published online Jun 19, 2007;
DOI: 10.1177/0146167207303022
The online version of this article can be found at:

http://psp.sagepub.com/cgi/content/abstract/33/8/1167
Published by:
http://www.sagepublications.com
On behalf of:
Society for Personality and Social Psychology, Inc.
Additional services and information for Personality and Social Psychology Bulletin can be found at:
Email Alerts: http://psp.sagepub.com/cgi/alerts
Subscriptions: http://psp.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Citations (this article cites 50 articles hosted on the

SAGE Journals Online and HighWire Press platforms):
http://psp.sagepub.com/cgi/content/refs/33/8/1167
Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

© 2007 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
Why Is Muscularity Sexy? Tests of the
Fitness Indicator Hypothesis
David A. Frederick
Martie G. Haselton
University of California, Los Angeles
Evolutionary scientists propose that exaggerated sec- commitment, and less so by physical attractiveness (e.g.,
ondary sexual characteristics are cues of genes that Buss & Schmitt, 1993; Pawlowski & Dunbar, 1999).
increase offspring viability or reproductive success. In six In the literature on nonhuman animals, however, there
studies the hypothesis that muscularity is one such cue is is much research on male attractiveness and far less on
tested. As predicted, women rate muscular men as sexier, female attractiveness (e.g., Alcock, 2005; Andersson,
more physically dominant and volatile, and less commit- 1994). Indeed, across species, females tend to be the sex
ted to their mates than nonmuscular men. Consistent that invests more in offspring and therefore they are more
with the inverted-U hypothesis of masculine traits, men selective in choosing mates (Trivers, 1972). Females
with moderate muscularity are rated most attractive. appear to value male attractiveness because it is a cue of
Consistent with past research on fitness cues, across two genes that confer fitness benefits to offspring through
measures, women indicate that their most recent short- increased viability or reproductive success (e.g., Kokko,
term sex partners were more muscular than their other Brooks, Jennions, & Morley, 2003; Moller, 1997). If spe-
sex partners (ds = .36, .47). Across three studies, when cific cues index heritable quality, females should come to
controlling for other characteristics (e.g., body fat), mus- value them in mates, and they should exercise this prefer-
cular men rate their bodies as sexier to women (partial rs = ence in mate selection.
.49-.62) and report more lifetime sex partners (partial Might male attractiveness also play heavily in the mat-
rs = .20-.27), short-term partners (partial rs = .25-.28), ing decisions of human females? Recent studies suggest
and more affairs with mated women (partial r = .28). that women discriminate between men on the basis of
hypothesized fitness cues, including facial masculinity
Keywords: body image; evolutionary psychology; mate pref-
erences; muscularity; sexual selection Authors’ Note: The authors are grateful to the University of
California, Los Angeles (UCLA) Graduate Division; the FPR-UCLA
Center for Culture, Brain, and Development; and the Departments of
P sychological research on physical attractiveness has Psychology and Communication Studies for providing financial sup-
largely focused on the importance of female attrac- port to the first author. For additional papers on body image and
body type preferences, please refer to the Web site of the first author
tiveness to men and the aspects of the female body men (http://dfred.bol.ucla.edu). We would like to thank Clark Barrett,
find most desirable in mates (e.g., Fink, Grammer, & Daniel Fessler, Gordon Gallup, Andrew Galperin, Kristina Grigorian,
Thornhill, 2001; Scutt, Manning, Whitehouse, Leinster, Kelsey Laird, Andrea Niles, Joshua Poore, Taylor Rhoades, Letitia
& Massey, 1997; Singh, 1993; Symons, 1995). Anne Peplau, Steven Platek, Leila Sadeghi-Azar, and the UCLA
Evolutionary psychologists have generally concurred that Experimental Biological Anthropology Lab for their helpful com-
ments on this manuscript and project. Correspondence should be
men possess strong preferences for female beauty because addressed to David A. Frederick, 1285 Franz Hall, Department of
attractive attributes are cues of fertility, and fertility Psychology, Third Floor Mailroom, University of California, Los
varies considerably between women and within individ- Angeles, CA 90095-1563; e-mail: enderflies1@aol.com.
ual women over time (for a review, see Sugiyama, 2005). PSPB, Vol. 33 No. 8, August 2007 1167-1183
In contrast, researchers have contended that men’s desir- DOI: 10.1177/0146167207303022
ability as mates is determined by earning potential and © 2007 by the Society for Personality and Social Psychology, Inc.
1167

1168 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN
(e.g., Johnston, Hagel, Franklin, Fink, & Grammer, 2001; associated with effort allocated to mating (McIntyre et al.,
Penton-Voak & Perrett, 2000; Penton-Voak et al., 1999) 2006) as well as with greater size and muscle mass (Bhasin,
and body scents associated with symmetry (Gangestad & 2003). Effort allocated to developing and maintaining
Thornhill, 1998). We examined whether women find these attributes can reduce budget of effort available for
muscular male bodies sexy and, if so, whether this prefer- maintaining other attributes (e.g., immunocompetence,
ence may also be a product of sexual selection. somatic upkeep) and can increase other energy demands
(e.g., increased metabolism; Buchanan, Evans, Goldsmith,
Bryant, & Rowe, 2001). This view suggests that there is a
SEXUAL SELECTION AND BODY MORPHOLOGY wider array of costs beyond simply immunosuppression
that causes these traits to be honest signals of quality.
Some traits are fitness cues because they demonstrate In both the immunocompetence and the more gen-
that a male is in good condition. Life history theorists eral cost models, however, the prediction is the same:
think of organisms as entities that capture energy from Traits produced by high levels of testosterone are cues
the environment and then convert it to survival and of heritable fitness or good condition because they indi-
reproduction-enhancing activities, including by develop- cate that the male can afford to generate these costly
ing metabolically expensive physical features that are traits. Selection should have shaped a female preference
attractive to the opposite sex (for a review, see Kaplan & for these traits because, all else equal, males displaying
Gangestad, 2005). Because of differences in genetic them sire more viable offspring.1
makeup, combined with the challenges faced during
development, individuals differ in their ability to allocate
energy to generating costly traits that are attractive to the THE COSTLY SIGNALING HYPOTHESIS AND
other sex. Zahavi (1975) proposed that males who dis- PREFERENCES FOR MUSCULARITY
play traits that are costly to maintain (e.g., the peacock’s
tail) are attractive to females precisely because they are We propose that the metabolic expense and levels of
costly and thus demonstrate that the male is in good testosterone necessary to build and sustain muscle mass
enough condition to produce them. Females who mate make muscularity a fitness cue. Numerous studies indi-
with these males would pass on the attractive traits to cate that increased muscle strength is associated with
their offspring, increasing their viability or reproductive naturally occurring levels of testosterone, as well as
success, or both. with testosterone treatments in normal, hypogonadal,
adolescent, and older male patients (e.g., Storer et al.,
2003; Wang et al., 2000; for a review, see Bhasin,
SEXUAL SELECTION AND 2003). Thus, men who are more muscular are exposed
TESTOSTERONE-LINKED TRAITS to greater levels of testosterone than other men.
For women to gain some genetic benefit by mating
As an extension of Zahavi’s (1975) hypothesis, with muscular men, however, muscularity must be her-
Folstad and Karter (1992) introduced the immunocom- itable. Estimates of the heritability of traits associated
petence signaling hypothesis. This hypothesis suggests with muscularity indicate that extremity circumferences
that secondary sexual characteristics are reliable indica- (e.g., bicep circumference), static strength (e.g., how
tors of mate quality because the reproductive hormones much weight a person can hold in place), and explosive
required for their development, including testosterone, strength (e.g., vertical jump) range from 20% to 80%
suppress the immune system (e.g., Peters, 2000; depending on the given trait (Loos et al., 1997; Thomis,
Rantala, Vainikka, & Kortet, 2003). The expression of Beunen, Maes, et al., 1998; Thomis, Beunen, Van
testosterone-linked traits reveals that men are in good Leemputte, et al., 1998; Thomis et al., 1997). One twin
enough condition to withstand the deleterious effects of study assessing gains in strength across a 10-week train-
immunosuppression, and women who selected these ing period also found that the ability to add muscle
men as mates would have transmitted features associ- mass beyond one’s baseline degree of muscularity is her-
ated with good condition to their offspring. itable (Thomis, Beunen, Maes, et al., 1998). The finding
An alternative perspective suggests that testosterone- that there are underlying genetic differences related to
linked traits are costly signals for reasons other than muscle mass indicates that muscularity, along with the
immunocompetence (see Kaplan & Gangestad, 2005; suite of traits correlated with it, can be passed on to off-
Kokko et al., 2003). In this view, fit males benefit more spring. This provides offspring with the advantage of
than other males from devoting a greater share of their developing traits that are attractive to females, further
energy budget to mating effort (competing for mates, dis- enhancing the women’s reproductive success in later
playing attributes desired by mates). Higher testosterone is generations.

Frederick, Haselton / MUSCULARITY AS A FITNESS INDICATOR 1169
STRATEGIC PLURALISM AND WOMEN’S Gangestad, Thornhill, & Garver-Apgar, 2005). This is
PREFERENCE FOR FITNESS INDICATORS especially true for women whose primary mates lack
sexual attractiveness—the women who, in theory, have the
Effects of Mating Context most to gain from extrapair mating with men who display
costly fitness indicators (Gangestad et al., 2005; Haselton
The perspective described previously predicts that & Gangestad, 2006; Pillsworth & Haselton, 2006).
women should find muscular men sexually desirable. One prediction that follows from the dual-mating logic
However, if muscular men are sexually desirable but is that men who display cues of fitness should be chosen
less likely to commit to their partners, women’s attrac- most often as affair partners. Symmetry is a purported
tion to muscularity should differ depending on mating index of fitness (see Moller, 1997); therefore, Thornhill
context. According to strategic pluralism theory and Gangestad (1994) examined partner number in men
(Gangestad & Simpson, 2000), men have evolved to varying in symmetry. As predicted, more symmetrical men
pursue reproductive strategies that are contingent on their reported having a greater overall number of sex partners,
value on the mating market. More attractive men accrue more sexual affairs, and a greater number of sex partners
reproductive benefits from spending more time seeking who were themselves mated to other men at the time of
multiple mating partners and relatively less time invest- the affair. Hughes and Gallup (2003) found a similar pat-
ing in offspring. In contrast, the reproductive effort of tern in men with higher shoulder-to-hip ratios, a trait that
less attractive men, who do not have the same mating may be linked with testosterone. In sum, both theory and
opportunities, is better allocated to investing heavily in existing evidence suggest that women attend to cues of fit-
their mates and offspring and spending relatively less ness when selecting sex partners, particularly short-term
time seeking additional mates. mates and affair partners.
From a woman’s perspective, the ideal is to attract a
partner who confers both long-term investment benefits
and genetic benefits. Not all women, however, will be THE INVERTED-U HYPOTHESIS OF
able to attract long-term investing mates who also dis- MASCULINE TRAITS
play heritable fitness cues. Consequently, women face
trade-offs in choosing mates because they may be forced When individuals consider others as mates, is more of
to choose between males displaying fitness indicators or a valued trait always better? Recent work by Kenrick and
those who will assist in offspring care and be good long- colleagues provided compelling evidence that the answer
term mates (Gangestad & Simpson, 2000). The most is no. For example, there comes a point where possessing
straightforward prediction that follows is that women additional income does not make one significantly more
seeking short-term mates, when the man’s only contri- desirable as a mate (Kenrick, Sundie, Nicastle, & Stone,
bution to offspring is genetic, should prefer muscularity 2001). Having more money, however, does not decrease
more than women seeking long-term mates. one’s attractiveness on the mating market.
In contrast to financial resources, there is reason to
Preferences in Extrapair Mates believe that high levels of masculine physical features,
including extreme muscularity and facial masculinity, can
As a partial solution to the problem of trade-offs, decrease a man’s desirability as a mate (Dixson, Halliwell,
women may have evolved to pursue a dual-mating East, Wignarajah, & Anderson, 2003; Johnston et al.,
strategy by securing investment from a long-term mate 2001). In a study by Johnston et al. (2001), women rated
and obtaining genetic benefits from extrapair mates the behaviors and dispositions of men varying in facial
(Haselton & Gangestad, 2006). Several lines of evi- masculinity. Facial masculinity was positively correlated
dence support this proposal. Although estimates vary, with sexual desirability, but the relationship was nonlin-
the human extrapair paternity rate is approximately 2% ear, with the most masculine faces perceived as being
to 4% (for a review, see Anderson, 2006). Thus, a sub- somewhat less sexually desirable, trustworthy, and sensi-
stantial portion of men raise offspring who are not tive than less masculine faces. Highly masculine faces
genetically their own. Men also appear to possess anti- were also rated as being more dominant, volatile, selfish,
cuckoldry mechanisms that lead them to detect the and impulsive than somewhat less masculine faces.
degree of resemblance between babies’ faces and their Johnston et al. concluded that “the aesthetic preference
own and adjust their investment accordingly (Platek of human females could be viewed as an adaptive com-
et al., 2003). Last, women are most attracted to men promise between the positive attributes associated with
other than their primary mate when fertility is high higher-than-average testosterone (health cues) and the
within the ovulatory cycle (and thus the benefits of negative attributes associated with more extreme mas-
extrapair mating for genetic benefits are highest; culinization” (p. 262).

We term this proposal the inverted-U hypothesis of shape, with muscular men being more attractive and
masculine traits: Women will not prefer mates with desirable than nonmuscular and very muscular men.
extremely high and extremely low levels of masculinity Furthermore, we predicted that women would infer that
(e.g., muscularity, facial masculinity, shoulder-to-hip ratio, very muscular men would be more likely to be physi-
and chest-to-waist ratio). Very high levels will be viewed as cally dominant and volatile compared with less muscu-
unattractive because these men are viewed as volatile and lar men. Last, consistent with the mating trade-off
threatening, perhaps presenting a direct danger to the hypothesis (Gangestad & Simpson, 2000), we predicted
woman. Low levels will be viewed as unattractive because that women would rate muscular men as less committed
these men are viewed as weak and submissive. Men with to their partners. These predictions were tested in
moderate to high levels should be preferred most as mates. Studies 1 and 2 by examining women’s ratings of the
attractiveness of computer-generated images and sil-
houettes of men varying in level of muscularity.
HYPOTHESIS AND PREDICTIONS
Women’s Preference for Muscular
We hypothesized that women possess context-sensi- Short-Term Partners
tive preferences for muscularity owing in part to an
underlying evolved psychology shaped by sexual selec- If muscularity is a cue of fitness, it should be more
tion. To investigate this hypothesis, we tested the fol- important to women selecting a short-term mate, when
lowing predictions. the man’s only contribution to offspring might be genetic
(Gangestad & Simpson, 2000). Although women may
Attraction to Muscularity also desire muscularity in long-term mates, strategic plu-
ralism theory predicts that not all women will be able to
If muscularity is a cue of fitness, women should be secure attractive mates as long-term partners. We pre-
more attracted to muscular men than to nonmuscular dicted that women would report preferring a more mus-
men. Past research generally supports this prediction. In cular short-term partner than long-term partner. This
questionnaire-based studies, women in Western societies prediction was tested in Study 2. Furthermore, we pre-
indicated that men with muscularity or high waist-to- dicted that women would report that their recent short-
chest ratios were attractive (e.g., Dixson et al., 2003; term sex partners were more muscular than their other
Franzoi & Herzog, 1987; Li & Kenrick, 2006; Maisey, sex partners. This prediction was tested in Study 3.
Vale, Cornelissen, & Tovee, 1999; Swami & Tovee,
2005) but not if they were highly muscular (Dixson et al., Muscularity Associated With Male Partner Number
2003). Although there has been little cross-cultural and Self-Perceived Desirability
research on women’s preferences for muscularity, some
evidence in non-Western societies suggests that women If women prefer muscularity in short-term mates,
prefer men with powerful body builds (Cassidy, 1991; muscular men should be able to capitalize on this prefer-
Dixson et al., 2003; for an exception, see Swami & ence and successfully attract multiple sex partners. Thus,
Tovee, 2005). In parallel, men in societies spanning four muscular men should report more lifetime sex partners,
continents believe that women are attracted to men who brief sexual affairs, and affairs with mated women than
are more muscular than average (Taiwan: Yang, Gray, & less muscular men. Male muscularity should also be pos-
Pope, 2005; Samoa: Lipinski & Pope, 2002; Austria and itively associated with self-rated attractiveness to women.
France: Pope et al., 2000; Kenya: Campbell, Pope, & These predictions were tested in Studies 4, 5, and 6.
Filliault, 2005; Ghana and the Ukraine: Frederick et al.,
in press; and the United States: Frederick et al., in press;
Olivardia, Pope, Borowiecki, & Cohane, 2004). The STUDY 1: SOCIAL PROFILES OF SIX COMUPTER-
majority of these studies, however, presented women GENERATED IMAGES OF MEN
with crude hand-drawn silhouettes of men. In Study 1,
we tested whether women find muscular men more sexu- This study investigated whether women believe mus-
ally desirable than nonmuscular men when evaluating cular men are more sexually desirable, more physically
relatively realistic computer-generated stimuli of men. dominant, more volatile, and less committed to their
romantic partners than less muscular men. We predicted
The Inverted-U Hypothesis of Masculine Traits that women’s ratings of the physical attractiveness and
sexual desirability of muscular men would show an
We predicted that the extent to which women find inverted-U pattern, with nonmuscular and very muscu-
muscularity attractive would follow an inverted-U lar men being rated as less attractive than moderately

muscular men. These predictions were tested by exam- divided by hip length) were similar among built (1.26),
ining women’s ratings of six computer-generated toned (1.24), and slender (1.24) individuals; slightly
images of men. larger for the brawny individual (1.32); and slightly
smaller among typical (1.15) and chubby (1.13) indi-
Method viduals. Finally, in an attempt to remove ethnic and
racial cues, faces were covered with a small black oval
Participants. A total of 141 undergraduate women and the images were printed on a laser printer in black
with a mean age of 20.44 (SD = 3.59) from the and white to yield images with ambiguous skin color.
University of California, Los Angeles (UCLA), partici-
pated in exchange for extra credit as part of their Procedure. Participants were asked 10 questions about
psychology or communication studies course. each image and made all ratings using a 9-point Likert
scale (1 = not at all, 3 = a little, 5 = somewhat, 7 = very,
Stimuli. The stimuli were images of shirtless men cre- 9 = extremely). The question stem for each item was
ated using MyVirtualModel.com, a program that allows “How likely is it that this man . . .” The participants
manipulation of physical features (see Appendix A). rated his physical dominance (“is physically intimidating
Aside from muscularity and body weight, all features of to other males”), his commitment to his partner (“would
the models were held constant, and height was set at 6 ft remain sexually faithful to you” and “would be sensitive
0 in. The first dimension varied was defined as muscular to your emotional needs”), his volatility (“has a bad tem-
versus nonmuscular by the program. The second dimen- per” and “would be abusive”), and his sexual desirabil-
sion was defined as total body weight by the program: ity (“would be sexually exciting,” “would be a good
large (230 lb), medium (190 lb), or small (150 lb). The sexual partner,” and “would be able to satisfy your
program also offers a limited ability to control shoulder sexual desires”). The sexual desirability category also
and waist proportions, and we attempted to standardize included the item “How physically attractive is this
shoulder-to-hip ratio across images. man?” All Cronbach’s alphas for each category with
There were six images in total: brawny (large, mus- more than one item, for each of the six images, were
cular), built (medium, muscular), toned (small, muscu- greater than .70. We therefore computed the category
lar), slender (small, nonmuscular), typical (medium, means for each image.
nonmuscular), and chubby (large, nonmuscular).
Participants were not exposed to these labels, only the Results and Discussion
images. These levels were chosen because of their face
validity; they appeared to differ systematically in body To examine differences in how women rated each
fat and muscularity. The validity of this manipulation body for each dimension (sexual desirability, domi-
was tested by presenting the images to 21 judges who nance, commitment, and volatility), we conducted a
rated how muscular and how fat each of the images one-way ANOVA with body type (brawny, built, toned,
appeared using a 0-100 scale (0 = not at all, 25 = a little, slender, typical, chubby) as the independent variable
50 = somewhat, 75 = very, 100 = extremely). Planned followed by planned comparisons of each cell. The
comparisons conducted within the context of one-way means are summarized in Figure 1 and Table 1. Because
ANOVAs revealed the following patterns of results for our predictions pertained to differences between partic-
muscularity (brawny > built > toned > slender = typical = ular cell means, we report the results of planned com-
chubby) and for body fat (brawny = built = toned = parisons among women’s ratings of each image rather
slender < typical < chubby) using p < .001 as the signif- than the results of the omnibus tests. Because we con-
icance criterion. Thus, to participants, there appeared to ducted multiple pairwise comparisons, we used a con-
be four levels of muscularity at the same level of body servative alpha level of .001. All differences were
fat: slender (small, nonmuscular), toned (small, muscu- significant at this level unless otherwise noted.
lar), built (medium, muscular), and brawny (large, mus- In support of the inverted-U hypothesis, brawny and
cular). As a preliminary test of the inverted-U slender men were rated as less sexually desirable than
hypothesis, we were interested in whether women found built men. They were also rated as less desirable than
men with moderate muscularity (toned, built) to be toned men, although the difference between brawny
more sexually desirable than men with very low or very and toned was marginally significant (p = .004). In
high levels of muscularity (slender, brawny). support of the predictions, each of the muscular men
To test whether the images differed in shoulder-to- (toned, built, and brawny) was rated as more dominant
hip ratio, a graphics designer unaffiliated with the than each of the nonmuscular men (slender, typical, and
project measured the length of the shoulders and hips chubby). Among the muscular men, the brawny man
using a graphics program. The ratios (shoulder length was rated the most dominant and the toned man was

TABLE 1: Women’s Ratings of Men Varying in Muscularity and Body Fat in Study 1
Sexual Desirability Physical Dominance Commitment Volatility
M SD M SD M SD M SD
Brawny 6.37 1.61 8.01 1.45 4.02 1.51 6.04 1.59

Built 6.97 1.42 7.18 1.54 4.58 1.33 5.37 1.42
Toned 6.87 1.31 5.70 1.53 5.16 1.25 4.37 1.42
Slender 5.42 1.59 2.99 1.79 5.88 1.31 3.45 1.52
Typical 4.28 1.35 3.98 1.66 5.84 1.37 3.84 1.49
Chubby 2.95 1.32 3.94 2.01 5.90 1.62 3.86 1.73
NOTE: Ratings of six computer-generated images of men were made on a 9-point Likert scale (1 = not at all, 9 = extremely) in Study 1. Women
reported that muscular men were more sexually desirable than nonmuscular men and that moderately muscular men (built, toned) were most
desirable. Women also reported that muscular men are more physically dominant and volatile but less likely to be committed to their partners.
8
Does man possess this trait? (Mean)
1
Sexual Desirability Physical Dominance Commitment Volatility
Brawny Built Toned Slender Typical Chubby
Figure 1 Association of body type with women’s social ratings of men.

NOTE: Ratings of six computer-generated images of men were made on a 9-point Likert scale (1 = not at all, 9 = extremely) in Study 1. Women
reported that muscular men were more sexually desirable than nonmuscular men and that moderately muscular men (built, toned) were most
desirable. Women also reported that muscular men are more physically dominant and volatile but less likely to be committed to their partners.
rated as the least dominant. We observed the same pat- facial masculinity, women rated men with moderate
tern of results for ratings of volatility, except that the muscularity rather than low or high muscularity as most
ratings of the slender and toned men only differed at a attractive, perhaps because men with low muscularity are
marginally significant level (p = .004). In support of the believed to possess too little dominance and men with
mating trade-off hypothesis, each of the muscular men high muscularity are believed to exhibit too much.
(toned, built, and brawny) was rated as being less com-
mitted than each of the nonmuscular men (slender, typ-
ical, chubby), and brawny men were perceived as least STUDY 2: THE INVERTED-U HYPOTHESIS
likely to be committed (all ps < .001). OF MASCULINE TRAITS
This pattern of results suggests that much like facial
masculinity, increased muscularity is associated with Study 1 found some support for the prediction that
inferences that a man is more physically dominant, more preferences for muscularity follow an inverted-U pat-
volatile, and less committed to his partner. Also similar to tern. However, the ability to test this hypothesis was

limited by the fact that only four levels of muscularity Results and Discussion
were presented to participants. In this study we tested
the inverted-U hypothesis more directly by examining Inverted-U hypothesis of masculine traits. A within-
women’s preferences for muscularity across eight sil- subjects ANOVA with linear and curvilinear (quadratic)
houette drawings of men varying in muscularity. contrasts was conducted to examine how the men’s mus-
Additionally, we tested a prediction derivable from mating cularity on the MSM related to women’s ratings of their
trade-off theory: that women’s preferences for muscu- physical attractiveness. Because we conducted multiple
larity will be stronger when they consider men as short- comparisons, we set a conservative alpha level of .001. As
term rather than long-term mates. shown in Figure 2, there was a significant effect of mus-
cularity, indicating that women’s ratings differed across
the various levels of muscularity, F(7, 1995) = 218.65,
Method p < .001. Pairwise comparisons indicated that most
Participants. A total of 286 women from UCLA with images differed from each other at the p < .001 level,
a mean age of 18.79 (SD = 1.40) volunteered to com- although some comparisons were only marginally signif-
plete a brief survey at the end of a lower-division psy- icant at the p < .05 level (Images 1 vs. 5 and Images 3 vs.
chology class. 5). Images 1 and 7, however, did not differ (p = .595).
There was a weak, marginally significant linear effect of
Stimuli. To assess women’s preferences for male muscularity, F(1, 285) = 9.17, p = .003, partial η2 = .031.
body types, women completed the Muscle Silhouette Specifically, there was a modest tendency for increased
Measure (MSM; Frederick et al., in press; see Appendix muscularity to be associated with lower ratings of attrac-
B). The MSM presents eight silhouettes of men varying tiveness, which is explicable by the fact that the extremely
from slender and nonmuscular to slender and extremely muscular men (Images 7-8) were rated particularly unat-
muscular. The MSM has also been used to assess body tractive. In support of the inverted-U hypothesis of mas-
satisfaction in the United States, Ukraine, and Ghana, culine traits, however, women’s ratings of muscularity
and has 1-month test–retest reliabilities exceeding r = followed a strong curvilinear relationship F(1, 285) =
.70 for men’s ratings of their current and desired bodies 1,189.68, p < .001, partial η2 = .807. These findings sup-
(Frederick et al., in press). It has also been used to dis- port the inverted-U hypothesis that increased muscularity
criminate between levels of muscularity represented in is attractive up to a point, with high and low levels of
popular male-audience, female-audience, and body- muscularity being less appealing to women.
builder-audience magazines (Frederick, Fessler, &
Haselton, 2005). In Studies 3-6 of this article, a corre- Short-term versus long-term partner preferences.
sponding measure to the MSM was also used: the Fat Consistent with the prediction derived from mating
Silhouette Measure (FSM; see Appendix C). The FSM trade-off theory, women reported that the best short-
presents eight silhouettes of men varying from slender term partner for them (M = 4.88, SD = 1.10) was more
and nonmuscular to obese and nonmuscular. It was also muscular than the best long-term partner for them (M =
used to examine body satisfaction among men in the 4.39, SD = 1.02), t(279) = 9.15, p < .001, d = .46.
United States, Ukraine, and Ghana (Frederick et al., in
press). Men’s self-ratings of their current body on this STUDY 3: MUSCULARITY OF WOMEN’S
form are highly correlated with body mass index (BMI; PAST SEXUAL PARTNERS
r = .69), a rough estimate of body fat level calculated
from self-reported height and weight (Frederick et al., The findings of Study 1 show that women find mus-
in press). cular men sexually attractive but believe they are less
likely to be committed, and the findings of Study 2 sug-
Procedure. The images were labeled 1-8, and inter- gest women prefer less muscularity in a long-term part-
vals of .5 were included so participants could indicate ner. These findings are consistent with the mating
intermediate values if they felt their answer lay between trade-off hypothesis, which suggests that women should
two images (1, 1.5, 2, 2.5, etc.). Women were first have stronger preferences for men displaying cues to her-
asked, “Which man would make the best short-term itable fitness, particularly when considering men as short-
sexual partner/brief sexual affair for you?” and “Which term mates. In parallel, men possessing these cues may be
man would make the best long-term dating partner for more likely to pursue a short-term mating strategy. Thus,
you?” They were then asked, “How attractive do you we predicted that women’s past short-term sex partners
find each person above?” Women rated each image would be more muscular than their other (longer term)
using a 9-point Likert scale: (1 = not at all, 3 = a little, sex partners. We tested this prediction by asking women
5 = somewhat, 7 = very, 9 = extremely). to rate the muscularity and athleticism of both their most

“How long were you dating the person before you had
9 intercourse?” Participants specified their answers in
Mean Rating of Physical
8 terms of weeks, months, and years, and all responses

Attractiveness
7 were converted to length of time in weeks. They were

6 then asked, “Before the first time you had sex . . . “how
5 emotionally close were you to this person?” “how
4 romantic was this person?” and “how much did you
3 trust this person?” Responses were made on a 9-point
2
Likert scale (1 = not at all, 3 = a little, 5 = somewhat,
1
7 = very, 9 = extremely). Women were also asked,
1
8
“Compared to the average man your age . . . “how mus-
e
e
ag
ag
ag
ag
ag
ag
ag
ag
Im
Im
Im
Im
Im
Im
Im
Im
cular was this person?” and “how athletic was this
Muscularity person?” Responses were made on a 9-point Likert
scale (1 = much less, 3 = somewhat less, 5 = equally,
Figure 2 Association of men’s muscularity with women’s ratings of 7 = somewhat more, 9 = much more). Finally, they
physical attractiveness. recorded the muscularity and body fat levels of these
NOTE: Image 1 represents a nonmuscular man on the Muscle
Silhouette Measure (MSM), whereas Image 8 represents an extremely partners using the MSM and the FSM, respectively.
muscular man (Frederick et al., in press; see Appendix B). Women
rated the physical attractiveness of each man on a 9-point Likert scale
(1 = not at all, 9 = extremely). Consistent with the inverted-U hypoth- Results and Discussion
esis of masculine traits, women’s ratings of the physical attractiveness
of men varying in muscularity followed a strong curvilinear pattern. Results are presented in Table 2. In support of the pre-
diction, women reported that their short-term partners
recent sex partner and their most recent short-term were more athletic and more muscular than their other
sex partner. recent sex partners on both the Likert scale measure and
the MSM. Women also reported they dated their short-
Method term partners for less time than their other recent sex
partners before having sex (1 week vs. 12 weeks), thus
Participants. Participants were drawn from a set of validating the temporal distinction between short-term
470 women from UCLA who participated in a larger partners and “other” partners. Relative to reports for
study on personality in which the target items for Study other partners, women also trusted their short-term part-
3 were embedded. Participants received research credit ners less, felt less emotionally close to them, and reported
as part of a psychology or communication course, or they were less romantic.
were entered in three lotteries for $50 each. A subset of One consequence of our method was that it was
82 women who indicated having a past short-term sex biased against the hypothesis because some of the indi-
partner, as well as another sex partner, completed a dat- viduals in the “other recent sex partner” category may
ing behavior survey and were included in the current also have been short-term partners. The fact that
study (M age = 22.35, SD = 5.23). women reported dating the short-term partners for sig-
nificantly less time than their other recent partners gives
Stimuli. Women reported their past partner’s muscu- us confidence that we were primarily comparing short-
larity and body fat using the MSM and FSM described term to long-term partners. However, to confirm that
in Study 2. our results were robust, we conducted all of the analy-
ses after eliminating anyone who indicated that she
Procedure. Women were asked to report several dated her most recent sex partner for less than 1 week
pieces of information about their most recent sex part- before having intercourse with him (n = 15). This did
ner, followed by their most recent short-term sex partner, not change the pattern of significant results.
who was defined as a person “whom you have had a Our finding that women reported greater muscular-
brief sexual affair or one-night stand with.” Women ity in their short-term partners than in their other part-
whose most recent sex partner was also their most ners is consistent with the hypothesis that muscularity is
recent short-term sex partner reported information a cue of fitness. Furthermore, it is noteworthy that
about their second-most-recent sex partner. women felt less emotionally close to their short-term
For both partners, women were asked, “How old partners than to their long-term partners before sex.
were you the first time you had intercourse with this One interpretation of this finding is that women were
person?” They were then asked, “How long did you biased to recall short-term partners more negatively
know the person before you had intercourse?” and than other partners. A second interpretation, consistent

TABLE 2: Characteristics of Past Sex Partners Reported in Study 2
Short Term Most Recent % ST % MR

(ST) (MR) t p d > MR > ST
Muscle (MSM) 3.76 (1.54) 3.23 (1.37) 2.85 .006 .36 61 28

Fat (FSM) 3.16 (1.14) 3.28 (1.30) –0.61 .545 –.10 38 46
Muscular 6.40 (1.93) 5.51 (1.87) 3.43 .001 .47 59 24
Athletic 6.23 (1.98) 5.81 (1.96) 2.06 .001 .33 59 25
Long dated 0.90 (2.05) 11.67 (21.23) –4.63 .001 –.71 6 76
Long known 48.49 (103.40) 45.12 (79.35) 0.26 .800 .04 37 60
Age 19.71 (3.78) 19.96 (3.93) –0.94 .349 –.06 32 38
Romantic 4.00 (2.34) 5.90 (2.34) –5.80 .001 –.82 18 67
Trust 4.51 (2.32) 6.54 (2.09) –6.28 .001 –.92 18 68
Emotionally close 3.45 (2.43) 5.51 (2.35) –6.34 .001 –.86 17 71
NOTE: The first two categories report women’s descriptions of their most recent short-term sex partners and other most recent partners, fol-
lowed by tests of whether these partners differed from each other on traits such as muscularity. The second-to-last column indicates the per-
centage of women who reported that their short-term sex partner scored higher on a trait (e.g., muscularity) than their most recent partner. The
last column indicates the percentage of women who reported that their most recent sex partner scored higher on a trait than their other recent
partners. MSM = Muscle Silhouette Measure; FSM = Fat Silhouette Measure.
with the mating trade-off hypothesis, is that women (Canon PowerShot S410, 4.0 Megapixels). Photographs
were more willing to have short-term relations with were taken in the same location under standardized light-
muscular men without the requirement that they ing conditions against a plain blue background.
demonstrate characteristics particularly desired in long-
term mates (trustworthiness, emotional closeness, etc.), Procedure. Four judges (3 women and 1 man) blind
possibly because these men possessed physical indica- to the purpose of the study coded the photographs.
tors of genetic fitness. They were instructed to first look at all of the pho-
tographs to get a sense of the range of body types. They
then coded the men for muscularity and body fat using
STUDY 4: GREATER SEX PARTNER NUMBER a 9-point Likert scale (1 = much less than average, 3 =
AMONG MUSCULAR MEN less than average, 5 = average, 7 = more than average,
9 = much more than average). They then coded the
The previous three studies revealed that women find men’s muscularity and body fat using the MSM and
muscular men sexually desirable and that women’s FSM described in Study 2. The judges’ ratings were
short-term sex partners were more muscular than their averaged for each item because the alphas were greater
other sex partners. These findings are congruent with than .70.
the hypothesis that men with fitness cues are selected
more often as sex partners. Here we test a correspond- Results and Discussion
ing prediction about men’s reports of their own sexual
histories: that muscular men would report having more Consistent with the predictions, greater muscularity,
lifetime sex partners than nonmuscular men. as coded by independent judges, was associated with
greater lifetime partner number (see Table 3), though
Method this association was not statistically significant (p >
.05). However, when we controlled for two important
Participants. The participants in this study were first confounds, age and body fat level, the association
photographed and then completed a survey on relation- between muscularity and partner number was statisti-
ships and sexuality. Men in this study (N = 115) were cally significant (see Table 3). This pattern was obtained
asked the following question as part of their survey: “In across each measure of muscularity (Likert scale and
your life so far, how many different people have you MSM). This finding supports the prediction that mus-
had sexual intercourse with?” Sixteen of the partici- cular men have more mating opportunities. One limita-
pants declined to be photographed, did not answer sex tion of this study, however, was that the participants
partner question, or both. The final sample included 99 were fully clothed, which may have obscured true dif-
men with a mean age of 21.26 (SD = 2.35). ferences between men and perhaps reduced associations
between muscularity and partner number. Studies 5 and 6
Stimuli. All of the men posed for standing full-body partly addressed this limitation by asking men to self-rate
digital photographs with their hands placed at their sides their level of muscularity.

TABLE 3: Correlations Between Self-Reported Muscularity (MSM), Attractiveness, and Sexual History
Muscularity r p Partial r r Partial-2 r p
Study 4
Lifetime sex partners (Muscularity 1-9) .09 .362 .27 .008
Lifetime sex partners (MSM) .06 .553 .20 .046
Participants (N) 99 94
Study 5
Sexy to average women .54 .001 .58 .001 .49 .001
Sexy to very attractive women .69 .001 .62 .001 .55 .001
Number of lifetime sex partners .27 .003 .24 .008 .26 .005
Number of short-term sex partners .20 .034 .18 .058 .25 .009
Participants (N) 112-123 108-119 107-118
Study 6
Sexy body to women .60 .001 .62 .001
Number of lifetime sex partners .25 .063 .27 .052
Number of short-term sex partners .32 .018 .33 .015
Number of affairs with mated women .27 .044 .28 .037
Participants (N) 56 52
NOTE: “Muscularity r” refers to the first-order correlations with muscularity on the Muscle Silhouette Measure (MSM) and other variables. The
only exception is that muscularity was also assessed on a Likert scale ranging from 1 to 9 in Study 4 (1 = much less than average, 9 = much more
than average). In Study 4, body fat and muscularity were determined by codings of participant photographs made by judges. In Studies 5 and 6,
body fat and muscularity were self-reported by participants. “Partial r” refers to the correlation between muscularity and the other variables,
controlling for body fat level and age. In Study 5, “Partial-2 r” also includes self-esteem as a control.
Studies 5 and 6: Greater sex partner number and self- Stimuli. The stimuli for Studies 5 and 6 were the
rated attractiveness among muscular men. Study 4 found MSM and FSM described in Study 2.
that men who were more muscular reported more life-
time sex partners. However, our muscularity hypothesis Procedure. In Studies 5 and 6, men were asked to
also predicts that increased muscularity will specifically indicate their level of muscularity and body fat on the
be associated with increased opportunities for short-term silhouette scales. In both studies, they also completed
sexual affairs. In Study 5, we tested the prediction that items using a 9-point Likert scale (1 = not at all, 3 = a
muscular men would report having more lifetime sex little, 5 = somewhat, 7 = very, 9 = extremely) designed
partners and short-term sex partners than nonmuscular to assess how attractive they believed women found
men, as well as greater self-reported attractiveness to their bodies. In Study 5 the items were “How sexy is
women. In Study 6, we attempted to replicate these find- your body to the average woman?” and “How sexy is
ings and to test the prediction that muscular men would your body to very attractive women?” In Study 6 the
report having more affairs with women who already have item was “How sexy is your body to women?” Men
a primary mate. This prediction follows from the mating also reported on their past dating history. In Studies 5
trade-off hypothesis, which proposes that women will and 6 they were asked, “How many individuals have
seek extrapair mates displaying fitness cues if their long- you had sexual intercourse with?” and “How many
term partners lack such cues. brief sexual affairs or one-night stands have you had?”
In Study 6 they were also asked, “How many times have
Method you had sex with a woman who had a boyfriend or
husband at the time you had sex with her?” Finally, in
Participants. In Study 5, heterosexual male students Study 5, participants completed a measure of general
(N = 124) from UCLA with a mean age of 20.90 (SD = self-esteem (Rosenberg, 1965) using a 4-point Likert
2.98) completed measures of past sexual history in scale, where higher scores represent greater self-esteem.
exchange for extra credit as part of a psychology or
communication studies class. In Study 6, heterosexual Results and Discussion
college men (N = 56) from UCLA with a mean age of
21.80 (SD = 6.92) participated in exchange for credit as Results for both studies are summarized in Table 3. As
part of a psychology class. predicted, compared with less muscular men, muscular

men rated their bodies as sexier to women in each mate preferences, a college-aged population is an appro-
study. Similarly, muscular men reported more lifetime priate place to begin because members of this popula-
sex partners, more short-term sex partners, and more tion are actively seeking and evaluating mates. Future
affairs with mated women at significant or marginally research is needed to examine how well the body types
significant levels. We then conducted partial correla- represented in this study actually map onto the levels of
tions controlling for participants’ age and self-reported muscularity present in traditional hunter–gatherer soci-
body fat level on the FSM to better isolate the associa- eties. Furthermore, an interesting question to explore is
tion of muscularity to self-rated attractiveness and sex whether women’s preferences for muscularity are rela-
partner number. The pattern of results remained essen- tively fixed or variable in response to local variation in
tially unchanged, although the association between body types and ecological conditions.
muscularity and number of brief sexual affairs in Study A second limitation is that men’s level of muscularity
5 dropped to marginally significant (p = .058). In Study was reported by participants in Studies 5 and 6 and was
5, we also included self-esteem as a control to rule out not objectively measured by researchers. When muscu-
the possibility that general self-confidence was driving larity was coded by independent judges in Study 4,
the association between self-reported muscularity and rather than relying on participant self-reports, muscu-
sex partner number. Despite the fact that self-reported larity was significantly associated with lifetime sex part-
muscularity and self-esteem were positively related (r = ner number when controlling for age and body fat level.
.39, p < .001), when self-esteem was added as a control, This indicates that muscularity is a good predictor of
muscularity was still a significant predictor of past lifetime sex partner number regardless of whether it is
number of sex partners and self-rated attractiveness.2 self-reported or coded by independent judges. Future
research could better estimate the strength of the asso-
ciations between muscularity and partner number by
GENERAL DISCUSSION photographing men with their upper torsos exposed.
One strength of this study was that conclusions
The results across six studies support the hypothesis about preferences for different body types were based
that muscularity is a sexually selected fitness cue. Women on computer-generated images of men in Study 1. Using
rated muscular men as more sexually desirable than both computer-generated images minimizes confounds asso-
nonmuscular men and very muscular men, as predicted ciated with images of real men and offers the ability to
by our inverted-U hypothesis of masculine traits (Studies systematically vary levels of muscularity. These images
1 and 2). Also consistent with our predictions, women were also more realistic than the hand-drawn silhou-
inferred that muscular men were more physically domi- ettes commonly used in studies of preferences for male
nant, more volatile, and less likely to show commitment body types (e.g., Campbell et al., 2005; Dixson et al.,
than less muscular men, a pattern consistent with 2003; Fallon & Rozin, 1985; Frederick et al., in press;
research on other testosterone-linked fitness indicators Lipinksi & Pope, 2002; Pope et al., 2000; Yang et al.,
such as facial masculinity (Johnston et al., 2001). 2005). It is possible, however, that women express dif-
If muscularity is a cue of fitness, muscular men ferent preferences when evaluating computer-generated
should have more mating opportunities. In Studies 3-6, images than when evaluating real men. In our previous
we found evidence supporting this prediction. Women research we asked women to rate unaltered pho-
reported that their past short-term sex partners were tographs of men varying in degrees of muscularity using
more muscular than their most recent sex partners methods similar to those outlined in Study 1 (Frederick,
across two measures of muscularity. Consistent with Haselton, Buchanan, & Gallup, 2003). Consistent with
past research on fitness cues (Hughes & Gallup, 2003; the findings in Study 1, greater muscularity was associ-
Thornhill & Gangestad, 1994), muscular men rated ated with ratings of more dominance, more volatility,
their bodies as sexier to women, reported more lifetime and less commitment, and ratings of sexual desirability
and short-term sex partners, and reported more affairs followed an inverted-U pattern. Thus, our results using
with mated women when controlling for other variables computer-generated images were cross-validated by
(age, body fat level, and self-esteem). results involving photographed men.
Finally, multiple methods were used to test the
Limitations and Strengths hypothesis that muscularity is a cue of fitness. For
example, women reported that muscular men were desir-
One limitation of these studies was that all of the able as sex partners and physically attractive. This was
participants were young college students who may paralleled by women’s reports that they had stronger
have preferences and mating patterns that differ from preferences for muscularity in a short-term sex partner
those of other populations. Nonetheless, in examining than in a long-term partner and that their short-term

partners were more muscular than their other partners. evolutionary theory of mate choice. This theory states
That women actually do choose muscular short-term that women prefer men with exaggerated secondary
partners was supported by the finding that muscular men sexual characteristics because these traits are cues of fit-
reported having more sex partners, brief sexual affairs, ness; therefore, women who were attracted to these men
and affairs with mated women. These pieces of evidence in our ancestral past would have had greater reproduc-
converge to indicate that muscularity is sexually attractive success than other women. However, unlike
tive and is possibly a cue of fitness. hypothesized fitness cues that only have value as a
social cue (e.g., facial masculinity), other evolutionary
The Evolution of Social Prestige and forces may have played a major role in shaping
Body Type Preferences women’s preference for muscularity. Muscularity itself
is likely to be directly useful in activities such as
The preceding arguments suggest that women pos- resource gathering and providing protection. Therefore,
sess evolved preferences specifically for muscularity. the preference for muscularity may have evolved in
One impressive feature of human psychology, of course, women, in part, because of the direct benefits muscular
is that humans appear to have evolved adaptations for men provided to their mates. Although we have focused
attending to, communicating, and processing socially on the genetic benefits underlying women’s preferences
and culturally transmitted information (e.g., Boyd & for muscularity, we do not rule out the influence of
Richerson, 2005). A competing perspective to the one sexual selection for direct benefits. Indeed, we concur
we have articulated would suggest that these prefer- with Wong and Candolin (2005), who observe,
ences are shaped solely or primarily by culturally trans-
mitted information about what body types are desirable Mate choice can present choosy individuals with both
(e.g., through the popular media). Although this is pos- direct material gains that increase their fecundity and/or
sible, in our view this competing explanation begs the survival, as well as indirect benefits that improve off-
spring viability and/or attractiveness. Competitive abil-
question of why certain body types become the subject
ity may correlate with some of these benefits if, for
of culturally transmitted information. Furthermore, it is example, males that are adept in competition also
not clear that these competing explanations would have monopolise the best resources or territories. . . .
predicted our specific pattern of findings. For example, Moreover, dominance could correlate with genetic ben-
whereas the mating trade-off theory (Gangestad & efits if sons inherit their father’s competitive prowess,
Simpson, 2000) explicitly predicts that women have dif- resulting in dominant males siring successful sons. (p. 2)
ferent preferences for short-term and long-term mates,
it is not clear that theories focusing solely on culturally Like dominance, muscularity could be preferred
transmitted values would have predicted this effect. because it is associated with both indirect (genetic) and
Future research is needed to fully test between these direct mating benefits.
competing models. One view that incorporates aspects of
both of these models has recently been advanced to explain Genetic or direct benefits? Examining the contexts
why muscular bodies are featured in the popular media. that accentuate preferences for muscularity could prove
Frederick et al. (2005) proposed that selection attached useful in determining whether women’s preferences are
heightened importance to some aspects of bodies because in fact due, at least in part, to genetic benefits. The strat-
of their association with fertility or virility, including mus- egy employed in this article was to examine whether
cularity. This enhances the likelihood that these aspects muscular men were chosen more often than less muscu-
will be featured in the media of cultural transmission (e.g., lar men as short-term sex partners and for affairs. It is
magazines, television, and movies) and seized on as not clear that a direct benefits perspective would predict
avenues for prestige competition (see Frederick et al., this association; thus, we view the data as providing
2005). More generally, future research should examine some support for the fitness indicator hypothesis.
how evolutionarily relevant factors might shift the impor- A further test of the genetic versus direct benefits
tance of muscularity to women and their preferred level of proposals would be to examine changes in women’s
muscularity (e.g., ecological factors, individual differences, preferences for muscularity across the ovulatory cycle.
parasite prevalence, degree of warfare). Past research has found that women’s preferences for
fitness indicators such as facial masculinity peak during
Other Routes to the Evolution of the high-fertility phase of the menstrual cycle, presum-
Preferences for Muscularity ably because this is when women would benefit most
from mating with men who possess fitness indicators
Genetic benefits, direct benefits, and intrasexual (e.g., Johnston et al., 2001; Penton-Voak & Perrett,
competition? These studies were motivated by a specific 2000; Penton-Voak et al., 1999). If muscularity is a

sexually selected fitness cue, preferences for muscular- than muscularity (e.g., height) provide evidence that a
ity may also peak during the high-fertility phase of the person was able to invest resources in somatomorphic
ovulatory cycle. The direct benefits perspective makes development during adolescence by developing a large
no such prediction. In a recent study, Gangestad, body frame. Other physical traits, such as athleticism
Garver-Apgar, Simpson, and Cousins (2006) asked and degree of muscularity, provide information not
women to watch videotapes of men varying on several only about past ability to invest in somatomorphic
dimensions (symmetry, facial masculinity, muscularity, development but also one’s current ability to invest
etc.). The researchers found that women preferred muscu- energy in maintaining a metabolically expensive body
lar men as short-term partners more during high-fertility type over an extended period. This malleable property
than during low-fertility phases of the ovulatory cycle. of muscularity may make it a particularly valuable cue,
A limitation of this study is that it is unknown whether as it provides evidence of a continued ability by men to
the effect was driven by muscularity, per se, or another maintain good condition and to devote resources to
feature that might correlate with muscularity in natural maintaining body size.
populations (e.g., facial masculinity). Future work in
this area using more controlled stimuli that vary only Conclusion
in muscularity is needed to test fully this prediction.
Although past research has focused primarily on the
The Relative Value of Muscularity as a Cue factors that make women physically attractive to men,
these studies suggest that men’s physical features are
The studies examined in this article suggest that mus- related not only to women’s expressed preferences for
cularity is one cue women use to assess men’s desirabil- mates but also to men’s and women’s past mate choices
ity as mates. Muscularity may partially covary with and sex behaviors. From a sexual selection perspective,
other traits, such as facial masculinity, symmetry, intra- this makes sense—these traits may be cues of heritable
sexual competitiveness and dominance, shoulder-to-hip fitness and women who expressed these preferences
ratio, waist-to-hip ratio, height, self-confidence, pres- would have had greater reproductive success than
tige, and socioeconomic status. Thus, research assessing women who did not. These findings expand on a grow-
the relative value of muscularity in predicting men’s ing body of literature suggesting that traits that are
desirability as mates is also needed. For example, costly to develop are important components of male
among men with similar shoulder-to-hip ratios, does physical attractiveness. These findings also support the
increased muscle definition (e.g., in the arms, abdomen, conclusion that male physical attractiveness plays a pre-
chest, and legs) have additional value as a cue of genetic viously underappreciated role in women’s mate choices—
or direct benefits (e.g., hunting ability)? One possibility is and perhaps in how men compete with each other to
that traits that are potentially more stable in adulthood attract women’s attention.

APPENDIX A
Images Representing Men Varying in Body Fat and Muscularity, Used in Study 1
NOTE: The images were created using MyVirtualModel.com.

APPENDIX B
Images Representing Men Varying in Muscularity, Used in Studies 2-6
NOTE: The figures are from the Muscle Silhouette Measure (Frederick et al., in press).
APPENDIX C
Images Representing Men Varying in Body Fat Level, Used in Studies 3-6
NOTE: The figures are from the Fat Silhouette Measure (Frederick et al., in press).
NOTES z-scored muscularity (Muscle Silhouette Measure) variable as predictor

of sex partner number. This had the effect of recoding low muscularity
and high muscularity with high values, so that a significant effect would
1. Some theorists propose that fitness indicators should be associ-
indicate a curvilinear association. Also included as predictors were age,
ated with good health (e.g., Rhodes et al., 2001). It is possible that
body fat (Fat Silhouette Measure), and the linear effect of muscularity
hypothesized fitness indicators such as muscularity or other masculine
as control variables. In Studies 4-6, there were no significant curvilinear
features could be associated with good health, but a necessary associ-
effects of muscularity on lifetime sex partner number, brief sexual
ation does not follow from sexual selection theory. For example,
affairs, affairs with mated women, or self-rated attractiveness to women
when there is strong within-sex competition, it may be advantageous
(all ps > .05). One reason for the lack of association could be that there
for individuals with relatively large energy budgets to spend extrava-
are simply not enough extremely muscular men in our sample to detect
gantly in signaling to attract mates, even if this compromises their
a significant curvilinear relationship. On the other hand, although
long-term health or chances of survival. Individuals with smaller bud-
highly muscular men might be less attractive to women, they may be
gets, on the other hand, may benefit more by pursuing a strategy of
more dominant, volatile, persistent, confident, coercive, narcissistic, and
lower resource expenditure, leading them to have better long-term
so on, which would lead them to have more sex partners.
health than other individuals. Furthermore, some costly signals may
also be useful for intrasexual competitions and intimidating rivals
(e.g., large horns or body size); thus, it may benefit some males to
develop very expensive features even if it drives their long-term health
below that of the average male. For more discussion of these issues, REFERENCES
please refer to reviews of the literature (e.g., Getty, 2002, 2006;
Kaplan & Gangestad, 2005). Alcock, J. (2005). Animal behavior: An evolutionary approach.
2. One interesting point to consider is whether the association Sunderland, MA: Sinauer.
between muscularity and sex partner number might also be curvilinear. Anderson, K. G. (2006). How well does paternity confidence match
If women find men with moderate muscularity most desirable, as sug- actual paternity? Evidence from worldwide nonpaternity rates.
gested by the inverted-U hypothesis, might moderate levels of mus- Current Anthropology, 47, 435-461.
cularity be associated with the highest number of sex partners? Andersson, M. (1994). Sexual selection. Princeton, NJ: Princeton
We tested the curvilinear effect by entering the squared values of the University Press.

Bhasin, S. (2003). Regulation of body composition by androgens. Kenrick, D. T., Sundie, J. M., Nicastle, L. D., & Stone, G .O. (2001).
Journal of Endocrinological Investigation, 26, 814-822. Can one ever be too wealthy or too chaste? Searching for nonlin-
Boyd, R., & Richerson, P. J. (2005). The origin and evolution of cul- earities in mate judgment. Journal of Personality and Social
tures. Oxford, UK: Oxford University Press. Psychology, 80, 462-471.
Buchanan, K. L., Evans, M. R., Goldsmith, A. R., Bryant, D. M., & Kokko, H., Brooks, R., Jennions, M. D., & Morley, J. (2003). The
Rowe, L. V. (2001). Testosterone influences basal metabolic rate evolution of mate choice and mating biases. Proceedings of the
in male house sparrows: A new cost of dominance signaling. Royal Society of London B, 270, 653-664.
Proceedings of the Royal Society of London B, 268, 1337-1344. Li, N. P., & Kenrick, D. T. (2006). Sex similarities and differences in
Buss, D. M., & Schmitt, D. P. (1993). Sexual strategies theory: An preferences for short-term mates: What, whether, and why.
evolutionary perspective on human mating. Psychological Review, Journal of Personality and Social Psychology, 90, 468-489.
100, 204-232. Lipinski, J. P., & Pope, H. G., Jr. (2002). Body ideals in young
Campbell, B. C., Pope, H. G., & Filliault, S. (2005). Body image Samoan men: A comparison with men in North America and
among Ariaal men from Northern Kenya. Journal of Cross- Europe. International Journal of Men’s Health, 1, 163-171.
Cultural Psychology, 36, 371-379. Loos, R., Thomis, M., Maes, H. H., Beunen, G., Claessens, A. L.,
Cassidy, C. M. (1991). The good body: When big is better. Medical Derom, C., et al. (1997). Gender-specific regional changes in
Anthropology, 13, 181-213. genetic structure of muscularity in early adolescence. Journal of
Dixson, A. F., Halliwell, G., East, R., Wignarajah, P., & Anderson, M. Applied Physiology, 82, 1802-1810.
(2003). Masculine somatotype and hirsuteness as determinants of Maisey, D. S., Vale, E. L., Cornelissen, P. L., & Tovee, M. J. (1999).
sexual attractiveness to women. Archives of Sexual Behavior, 32, Characteristics of male attractiveness for women. Lancet, 353,
29-39. 1500.
Fallon, A. E., & Rozin, P. (1985). Sex differences in perceptions of desir- McIntyre, M., Gangestad, S. W., Gray, P. B., Chapman, J. F.,
able body shape. Journal of Abnormal Psychology, 94, 102-105. Burnham, T. C., O’Rourke, M. T., et al. (2006). Romantic
Fink, B., Grammer, K., & Thornhill, R. (2001). Human (Homo sapi- involvement often reduces men’s testosterone levels—but not
ens) facial attractiveness in relation to skin texture and color. always. The moderating role of extrapair sexual interest. Journal
Journal of Comparative Psychology, 115, 92-99. of Personality and Social Psychology, 91, 642-651.
Folstad, I., & Karter, A. (1992). Parasites, bright males, and Moller, A. P. (1997). Developmental stability and fitness: A review.
the immunocompetence handicap. American Naturalist, 139, American Naturalist, 149, 916-932.
603-622. Olivardia, R., Pope, H. G., Borowiecki, J. J., & Cohane, G. H.
Franzoi, S. L., & Herzog, M. E. (1987). Judging physical attractive- (2004). Biceps and body image: The relationship between muscu-
ness: What body aspects do we use? Personality & Social larity and self-esteem, depression, and eating disorder symptoms.
Psychology Bulletin, 13, 19-33. Psychology of Men and Masculinity, 5, 112-120.
Frederick, D. A., Buchanan, G. M., Berezovskaya, A., Peplau, L. A., Pawlowski, B., & Dunbar, R. I. (1999). Impact of market value on
Haselton, M. G., & Lipinski, R. E. (in press). Desiring the muscu- human mate choice decisions. Proceedings of the Royal Society of
lar ideal: Men’s body satisfaction in the United States, Ukraine, London B, 266, 281-285.
and Ghana. Psychology of Men and Masculinity. Penton-Voak, I. S., & Perrett, D. I. (2000). Female preference for male
Frederick, D. A., Fessler, D. M. T., & Haselton, M. G. (2005). Do rep- faces changes cyclically: Further evidence. Evolution and Human
resentations of male muscularity differ in men’s and women’s maga- Behavior, 21, 39-48.
zines? Body Image: An International Journal of Research, 2, 81-86. Penton-Voak, I. S., Perrett, D. I., Castles, D. L., Kobayashi, T., Burt,
Frederick, D. A., Haselton, G. M., Buchanan, G. M., & Gallup, G. G., D. M., Murray, L. K., et al. (1999). Menstrual cycle alters face
Jr. (2003). An evolved preference for muscularity? Evidence from preference. Nature, 399, 741-742.
women’s preferences for short-term and long-term partners. Paper Peters, A. (2000). Testosterone treatment is immunosuppressive in
presented at the Human Evolution and Behavior Conference, superb fairy-wrens, yet free-living males with high testosterone are
Lincoln, NE. more immunocompetent. Proceedings of the Royal Society of
Gangestad, S. W., Garver-Apgar, C. E., Simpson, J. A., & Cousins, A. J. London B, 267, 883-889.
(2007). Changes in women’s mate preferences across the ovulatory Pillsworth, E. G., & Haselton, M. G. (2006). Male sexual attractive-
cycle. Journal of Personality and Social Psychology, 92, 151-163. ness predicts differential ovulatory shifts in female extra-pair
Gangestad, S. W., & Simpson, J. A. (2000). The evolution of human attraction and male mate retention. Evolution and Human
mating: Trade-offs and strategic pluralism. Behavioral & Brain Behavior, 27, 247-258.
Sciences, 23, 573-644. Platek, S. M., Critton, S. R., Burch, R. L., Frederick, D. A., Myers,
Gangestad, S. W., & Thornhill, R. (1998). Menstrual cycle variation T. E., & Gallup, G. G., Jr. (2003). How much paternal resem-
in women’s preference for the scent of symmetrical men. blance is enough? Sex differences in hypothetical investment deci-
Proceedings of the Royal Society of London B, 265, 927-933. sions but not in the detection of resemblance. Evolution and
Gangestad, S. W., Thornhill, R., & Garver-Apgar, C. E. (2005). Human Behavior, 24, 81-87.
Adaptations to ovulation. In D. Buss (Ed.), The handbook of evo- Pope, H. G., Gruber, A. J., Mangweth, B., Bureau, B., deCol, C.,
lutionary psychology (pp. 292-343). New York: John Wiley. Jouvent, R., et al. (2000). Body image perception among men in
Getty, T. (2002). Signalling health versus parasites. American three countries. American Journal of Psychiatry, 157, 1297-1301.
Naturalist, 159, 364-371. Rantala, M. J., Vainikka, A., & Kortet, R. (2003). The role of juve-
Getty, T. (2006). Sexually selected signals are not similar to sports nile hormone in immune function and pheromone production
handicaps. Trends in Ecology and Evolution, 21, 83-88. trade-offs: A test of the immunocompetence handicap principle.
Haselton, M. G., & Gangestad, S. W. (2006). Conditional expression Proceedings of the Royal Society of London B, 270, 2257-2261.
of women’s desires and men’s mate guarding across the ovulatory Rhodes, G., Zebrowitz, L. A., Clark, A., Kalick, S. M., Hightower, A.,
cycle. Hormones and Behavior, 49, 509-518. & McKay, R. (2001). Do facial averageness and symmetry signal
Hughes, S. M., & Gallup, G. G., Jr. (2003). Sex differences in mor- health? Evolution and Human Behavior, 22, 31-46.
phological predictors of sexual behaviors: Shoulder to hip and Rosenberg, M. (1965). Society and the adolescent self-image.
waist to hip ratios. Evolution and Human Behavior, 24, 173-178. Princeton, NJ: Princeton University Press.
Johnston, V. S., Hagel, R., Franklin, M., Fink, B., & Grammer, K. Scutt, D., Manning, J. T., Whitehouse, G. H., Leinster, S. J., &
(2001). Male facial attractiveness. Evidence for hormone-medi- Massey, C. P. (1997). The relationship between breast asymmetry,
ated adaptive design. Evolution and Human Behavior, 22, 251- breast size and the occurrence of breast cancer. British Journal of
267. Radiology, 70, 1017-1021.
Kaplan, H. S., & Gangestad, S. W. (2005). Life history theory and Singh, D. (1993). Adaptive significance of female physical attractive-
evolutionary psychology. In D. Buss (Ed.), The handbook of evo- ness: Role of waist-to-hip ratio. Journal of Personality and Social
lutionary psychology (pp. 68-95). New York: John Wiley. Psychology, 65, 293-307.

Storer, T. W., Magliano, L., Woodhouse, L., Lee, M. L., Dzekov, C., analysis of maximal isometric force at different elbow angles.
Dzekov, J., et al. (2003). Testosterone dose-dependently increases Journal of Applied Physiology, 82, 959-967.
maximal voluntary strength and leg power, but does not affect Thornhill, R., & Gangestad, S. W. (1994). Human fluctuating asym-
fatigability or specific tension. Journal of Clinical Endocrinology metry and sexual behavior. Psychological Science, 5, 297-302.
& Metabolism, 88, 1478-1485. Trivers, R. (1972). Parental investment and sexual selection. In B.
Sugiyama, L. S. (2005). Physical attractiveness in adaptationist per- Campbell (Ed.), Sexual selection and the descent of man: 1871-
spective. In D. M. Buss (Ed.), The handbook of evolutionary psy- 1971 (pp. 136-179). Chicago: Aldine.
chology (pp. 292-343). New York: John Wiley. Wang, C., Wang, C., Swerdloff, R. S., Iranmanesh, A., Dobs, A.,
Swami, V., & Tovee, M. J. (2005). Male physical attractiveness in Snyder, P. J., et al. (2000). Transdermal testosterone gel improves
Britain and Malaysia: A cross-cultural study. Body Image: An sexual function, mood, muscle strength, and body composition
International Journal of Research, 2, 383-393. parameters in hypogonadal men. American Journal of Physiology,
Symons, D. (1995). Beauty is in the adaptations of the beholder: The Endocrinology, and Metabolism, 85, 2839-2853.
evolutionary psychology of human female sexual attractiveness. In Wong, B. B. M., & Candolin, U. (2005). How is female mate choice
P. R. Abramson & S. D. Pinkerton (Eds.), Sexual nature, sexual affected by male competition. Biological Reviews of the
culture (pp. 80-119). Chicago: University of Chicago Press. Cambridge Philosophical Society, 80, 559-571.
Thomis, M. A., Beunen, G. P., Maes, H. H., Cameron, J., Van Yang, C. F. J., Gray, P., & Pope, H. G., Jr. (2005). Male body image
Leemputte, M., Claessens, A. L., et al. (1998). Strength training: in Taiwan versus the West: Yangghang Zhiqi meets the Adonis
Importance of genetic factors. Medicine & Science in Sports & Complex. American Journal of Psychiatry, 162, 263-269.
Exercise, 30, 724-731. Zahavi, A. (1975). Mate selection—A selection for a handicap.
Thomis, M. A., Beunen, G. P., Van Leemputte, M., Maes, H. H., Journal of Theoretical Biology, 53, 205-214.
Blimkie, C. J., Claessen, A. L., et al. (1998). Inheritance of static
and dynamic arm strength and some of its determinants. Acta Received June 19, 2006
Physiologica Scandinavica, 163, 59-71. Revision accepted February 1, 2007
Thomis, M. A., Van Leemputte, M. V., Maes, H. H., Blimkie, C. J.,
Claessens, A. L., Marchal, G., et al. (1997). Multivariate genetic


Frederick Haselton+2007

Uploaded by

Copyright:

Available Formats

Frederick Haselton+2007

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Frederick Haselton+2007

Uploaded by

Copyright:

Available Formats

Personality and Social Psychology Bulletin

Why Is Muscularity Sexy? Tests of the Fitness Indicator Hypothesis

The online version of this article can be found at:

Society for Personality and Social Psychology, Inc.

Email Alerts: http://psp.sagepub.com/cgi/alerts

Citations (this article cites 50 articles hosted on the

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Sexual Desirability Physical Dominance Commitment Volatility

Brawny 6.37 1.61 8.01 1.45 4.02 1.51 6.04 1.59

Brawny Built Toned Slender Typical Chubby

Figure 1 Association of body type with women’s social ratings of men.

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

8 terms of weeks, months, and years, and all responses

7 were converted to length of time in weeks. They were

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

TABLE 2: Characteristics of Past Sex Partners Reported in Study 2

Short Term Most Recent % ST % MR

Muscle (MSM) 3.76 (1.54) 3.23 (1.37) 2.85 .006 .36 61 28

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Muscularity r p Partial r r Partial-2 r p

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

NOTE: The images were created using MyVirtualModel.com.

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Images Representing Men Varying in Muscularity, Used in Studies 2-6

NOTES z-scored muscularity (Muscle Silhouette Measure) variable as predictor

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

Downloaded from http://psp.sagepub.com at PENNSYLVANIA STATE UNIV on April 11, 2008

You might also like