diff --git a/doc/usage.rst b/doc/usage.rst
index 95e9c25d..eba41005 100644
--- a/doc/usage.rst
+++ b/doc/usage.rst
@@ -185,7 +185,7 @@ options::
   -v, --verbose         Print more output
   -O STYLE, --output_style STYLE
                         What style the benchmark output should take.
-                        Valid options are 'normal' and 'table'.
+                        Valid options are 'normal', 'table', and 'table_github'.
                         Default is normal.
   --csv CSV_FILE        Name of a file the results will be written to,
                         as a three-column CSV file containing minimum
diff --git a/pyperformance/cli.py b/pyperformance/cli.py
index df68dc3d..bde9a9eb 100644
--- a/pyperformance/cli.py
+++ b/pyperformance/cli.py
@@ -113,10 +113,10 @@ def parse_args():
     cmd.add_argument("-v", "--verbose", action="store_true",
                      help="Print more output")
     cmd.add_argument("-O", "--output_style", metavar="STYLE",
-                     choices=("normal", "table"),
+                     choices=("normal", "table", "table_github"),
                      default="normal",
                      help=("What style the benchmark output should take."
-                           " Valid options are 'normal' and 'table'."
+                           " Valid options are 'normal', 'table', and 'table_github'."
                            " Default is normal."))
     cmd.add_argument("--csv", metavar="CSV_FILE",
                      action="store", default=None,
diff --git a/pyperformance/compare.py b/pyperformance/compare.py
index 1302733c..e3cf7a05 100644
--- a/pyperformance/compare.py
+++ b/pyperformance/compare.py
@@ -202,6 +202,29 @@ def format_table(base_label, changed_label, results):
     output.insert(2, "".join(header_sep_line))
     return "\n".join(output)
 
+def _format_github_row(items):
+    return "| " + " | ".join(items) + " |"
+
+def format_github_table(base_label, changed_label, results):
+    columns = ("Benchmark", base_label, changed_label, "Change", "Significance")
+    output = [_format_github_row(columns), "| --- " * len(columns) + "|"]
+
+    for (bench_name, result) in results:
+        format_value = result.base.format_value
+        avg_base = result.base.mean()
+        avg_changed = result.changed.mean()
+        delta_avg = quantity_delta(result.base, result.changed)
+        msg = significant_msg(result.base, result.changed)
+        rows = (bench_name,
+                # Limit the precision for conciseness in the table.
+                format_value(avg_base),
+                format_value(avg_changed),
+                delta_avg,
+                msg)
+        output.append(_format_github_row(rows))
+
+    return "\n".join(output)
+
 
 class BenchmarkResult(object):
     """An object representing data from a succesful benchmark run."""
@@ -355,6 +378,9 @@ def compare_results(options):
     elif options.output_style == "table":
         if shown:
             print(format_table(base_label, changed_label, shown))
+    elif options.output_style == "table_github":
+        if shown:
+            print(format_github_table(base_label, changed_label, shown))
     else:
         raise ValueError("Invalid output_style: %r" % options.output_style)