diff --git a/contrib/perf.py b/contrib/perf.py
--- a/contrib/perf.py
+++ b/contrib/perf.py
@@ -15,6 +15,16 @@ Configurations
 ``presleep``
   number of second to wait before any group of run (default: 1)
 
+``run-limits``
+  Control the number of run each benchmark will perform. The option value
+  should be a list of `<time>-<numberofrun>` pairs. After each run the
+  condition are considered in order with the following logic:
+
+      If benchmark have been running for <time> seconds, and we have performed
+      <numberofrun> iterations, stop the benchmark,
+
+  The default value is: `3.0-100, 10.0-3`
+
 ``stub``
     When set, benchmark will only be run once, useful for testing (default: off)
 '''
@@ -225,6 +235,9 @@ try:
     configitem(b'perf', b'all-timing',
         default=mercurial.configitems.dynamicdefault,
     )
+    configitem(b'perf', b'run-limits',
+        default=mercurial.configitems.dynamicdefault,
+    )
 except (ImportError, AttributeError):
     pass
 
@@ -297,7 +310,34 @@ def gettimer(ui, opts=None):
 
     # experimental config: perf.all-timing
     displayall = ui.configbool(b"perf", b"all-timing", False)
-    return functools.partial(_timer, fm, displayall=displayall), fm
+
+    # experimental config: perf.run-limits
+    limitspec = ui.configlist(b"perf", b"run-limits", [])
+    limits = []
+    for item in limitspec:
+        parts = item.split('-', 1)
+        if len(parts) < 2:
+            ui.warn(('malformatted run limit entry, missing "-": %s\n'
+                     % item))
+            continue
+        try:
+            time_limit = float(parts[0])
+        except ValueError as e:
+            ui.warn(('malformatted run limit entry, %s: %s\n'
+                     % (e, item)))
+            continue
+        try:
+            run_limit = int(parts[1])
+        except ValueError as e:
+            ui.warn(('malformatted run limit entry, %s: %s\n'
+                     % (e, item)))
+            continue
+        limits.append((time_limit, run_limit))
+    if not limits:
+        limits = DEFAULTLIMITS
+
+    t = functools.partial(_timer, fm, displayall=displayall, limits=limits)
+    return t, fm
 
 def stub_timer(fm, func, setup=None, title=None):
     if setup is not None:
diff --git a/tests/test-contrib-perf.t b/tests/test-contrib-perf.t
--- a/tests/test-contrib-perf.t
+++ b/tests/test-contrib-perf.t
@@ -55,6 +55,16 @@ perfstatus
   "presleep"
     number of second to wait before any group of run (default: 1)
   
+  "run-limits"
+    Control the number of run each benchmark will perform. The option value
+    should be a list of '<time>-<numberofrun>' pairs. After each run the
+    condition are considered in order with the following logic:
+  
+        If benchmark have been running for <time> seconds, and we have performed
+        <numberofrun> iterations, stop the benchmark,
+  
+    The default value is: '3.0-100, 10.0-3'
+  
   "stub"
       When set, benchmark will only be run once, useful for testing (default:
       off)
@@ -231,6 +241,31 @@ perfstatus
   $ hg perfparents
   $ hg perfdiscovery -q .
 
+Test run control
+----------------
+
+Simple single entry
+
+  $ hg perfparents --config perf.stub=no --config perf.run-limits='0.000000001-15'
+  ! wall * comb * user * sys * (best of 15) (glob)
+
+Multiple entries
+
+  $ hg perfparents --config perf.stub=no --config perf.run-limits='500000-1, 0.000000001-5'
+  ! wall * comb * user * sys * (best of 5) (glob)
+
+error case are ignored
+
+  $ hg perfparents --config perf.stub=no --config perf.run-limits='500, 0.000000001-5'
+  malformatted run limit entry, missing "-": 500
+  ! wall * comb * user * sys * (best of 5) (glob)
+  $ hg perfparents --config perf.stub=no --config perf.run-limits='aaa-12, 0.000000001-5'
+  malformatted run limit entry, could not convert string to float: aaa: aaa-12
+  ! wall * comb * user * sys * (best of 5) (glob)
+  $ hg perfparents --config perf.stub=no --config perf.run-limits='12-aaaaaa, 0.000000001-5'
+  malformatted run limit entry, invalid literal for int() with base 10: 'aaaaaa': 12-aaaaaa
+  ! wall * comb * user * sys * (best of 5) (glob)
+
 test actual output
 ------------------