Enter one honest set and this rates it against the reference table most gyms and textbooks use. Read the limits underneath before you take the number anywhere: they matter more than the number does.
Where these numbers come from, and what they are not
The thresholds are the adult push-up norms from the Canadian Society for Exercise Physiology's CSEP-PATH protocol, the table reproduced in most fitness textbooks and by the American College of Sports Medicine. Four things about them are worth knowing before you quote your result to anyone.
They are category boundaries, not measured percentiles. CSEP publishes five rating bands. The percentile figures here are interpolated between those bands, which is why the calculator rounds to the nearest five and refuses to claim more precision than that. Nobody measured a 63rd percentile.
The men's and women's numbers are not comparable. In the CSEP protocol adult men perform standard push-ups and adult women perform them from the knees. These are two different exercises, so a man's 25 and a woman's 25 are not the same achievement in either direction, and the calculator will not convert between them.
The reference test is paced, and yours probably was not. CSEP's push-up test runs to a metronome at 20 push-ups per minute, one every three seconds, continuing until form fails. A maximum set done at your own speed is a different measurement, and usually a more generous one, because going fast lets you use bounce that the paced protocol removes. If you want a number that genuinely compares, set a three second cadence and stop when your form breaks rather than when your arms do.
The sample is small and old. The adult table traces to a Canadian dataset of a few hundred people, validated against a larger national survey, published in 2000. It is the best widely used reference available, which is not the same as being a definitive picture of how strong people are in 2026.
What a percentile is actually good for
Very little, on its own. It tells you where one set on one day sits against a small Canadian sample from a quarter of a century ago. It tells you nothing about whether you are getting stronger, which is the only question that matters for training.
The useful version is the same test repeated. Do it under the same conditions, note the number, and repeat it in eight weeks. A move from 18 to 24 is real information about you. A move from the 55th to the 70th percentile is the same information wearing a costume.
If your set was not a fair test
- You stopped when it got hard, not when form broke. The reference test ends at form failure, so a set you cut short scores low for reasons that have nothing to do with strength.
- You went fast. Speed inflates the count against a protocol built around a three second cadence.
- You did not reach depth. The test requires the elbows to bend to about ninety degrees. Half reps are not reps, on this table or anywhere else.
- You were not fresh. A max test after a workout measures your fatigue, not your ceiling.
Our guide to testing your max push-ups properly covers the conditions that make a number worth writing down.
Where Pushup RPG fits, and our bias
We make Pushup RPG, so treat this paragraph accordingly. The app counts push-ups with the phone camera and only credits a rep that reaches depth, which removes the most common way a self-reported max is wrong. It does not pace you at three seconds a rep, so a number from the app is still a maximum-effort count rather than a CSEP test result. If you want to run the reference protocol properly, use a metronome and count the honest way, whether or not any app is involved.
