Between 27 and 49 reps in one minute for a man under 45, depending on your age, if you want a number that came from a real sample. The YMCA's one-minute sit-up norms — published by Golding and colleagues in The Y's Way to Physical Fitness — rate an average man aged 18-25 at 35-38 reps, good at 44-49, and excellent above 49. For a man aged 36-45, average is 27-29 and excellent is above 41. Which makes the chart circulating on page one of Google a problem: it tells a 30-something man to aim for 40, calls that a standard, and traces back to a single trainer quoted in a news article. Here is the other half, and it matters more than the number. The Navy, the Army and the Marine Corps have all deleted the sit-up from their fitness tests since 2020. Score yourself if you want a benchmark. Just don't build your ab training on a test the people who invented mass fitness testing threw out.
How many sit-ups can the average man do in a minute?
Thirty-five to 38 reps for a man aged 18-25. That is the "average" band in the only widely cited norm table with an actual testing program behind it.
The source almost nobody names is Lawrence Golding, whose The Y's Way to Physical Fitness (3rd edition, 1986, pp. 113-124) carries the YMCA fitness battery's norms. Every sit-up chart you have ever seen is downstream of it, usually without the citation. The full male 18-25 column, in reps performed in 60 seconds:
- Excellent — above 49
- Good — 44 to 49
- Above average — 39 to 43
- Average — 35 to 38
- Below average — 31 to 34
- Poor — 25 to 30
- Very poor — under 25
The decay across the decades is real and steep. Men 36-45: excellent above 41, good 35-41, average 27-29, poor 17-22. Men 65 and over: excellent above 28, good 22-28, average 15-18, poor 7-10. Women 18-25 run excellent above 43, good 37-43, average 29-32; women 65+ average 11-13.
Now the honest label, because this table gets quoted as if it were a census. It is forty years old, and the people tested were YMCA members — adults who had already paid to join a gym, which is not a random slice of anything. The publisher puts the norms at roughly 20,000 testees across six age groups per sex; I could not open the 1986 book to check that count, so treat it as the publisher's claim. Read the table as the best available benchmark, not the truth about all men. It still beats everything else on this keyword for one reason: somebody actually counted.
Is there an official sit-up standard?
Two, and they disagree — because they are grading two different movements.
The first is the US Army Physical Fitness Test, the Army's test of record for roughly four decades until the Army Combat Fitness Test took over. Its sit-up event is spelled out in the Army's own Training Circular TC 3-22.20, 20 August 2010, Appendix A: two minutes, knees bent at 90 degrees, another person holding your ankles by hand, fingers interlocked behind your head, up until the base of your neck passes the base of your spine, down until the bottoms of your shoulder blades touch the ground. The up position is the only legal rest.
One line in that document is worth more than any score table. The narrative read aloud to every soldier before the event says: "THE SIT-UP EVENT MEASURES THE ENDURANCE OF THE ABDOMINAL AND HIP-FLEXOR MUSCLES." The Army wrote down, in its own manual, that its ab test is partly a hip-flexor test. Hold that thought for two sections.
I am not printing APFT point-conversion numbers here, deliberately. Those tables live as scanned images inside DA Form 705, and every copy I could reach was either unreadable or a secondary site's retyping — including one top-ranking page whose table is visibly scrambled. A benchmark you can't trace isn't a benchmark.
The second standard is the partial curl-up used in the Canadian fitness battery and picked up by ACSM, built by Faulkner, Sprigings, McQuarrie and Bell, whose 262-adult protocol study led Fitness Canada to adopt it. It is not a sit-up. Arms rest at your sides, fingertips slide 10 cm along the floor, your feet are not held, and you move to a metronome at 50 beats per minute for one minute. Take the feet away and cap the range of motion and you remove most of the hip flexor — the whole point of the design, and why a partial curl-up score and a YMCA sit-up score are not interchangeable. Same body part, different test, different numbers.
Where does the sit-ups-by-age chart on page one come from?
One named trainer, quoted in one article, with no sample behind it — and a second, different trainer with different numbers in another.
UNILAD published the version that has spread furthest, in late November 2025, quoting Nico Schwandt, a physiologist at the Stone London gyms: 40 crunches for 30-39-year-olds, 30 at 40-49, 20 at 50-59, 10 at 60-69 and 5 at 70-plus. To its credit, the article does not oversell them: the numbers are described as "markers to aim for" rather than strict standards. Then the caveat gets stripped and the digits get syndicated.
Fox News ran its own crunches-by-age piece on 31 March 2025 with a completely different trainer — Joseph David of Life Time in New York — and a completely different table: 40-50 in your 20s, 30-40 in your 30s, 20-30 in your 40s, 15-25 in your 50s, 10-20 at 60-plus. No study, no sample. The most useful sentence in it comes from the second trainer quoted, Kollins Ezekh: "There's no magic number you 'should' be able to do based on age, since fitness levels really vary."
Set those against Golding. Page one tells a man in his thirties to hit 40 crunches. Golding's measured band for a man aged 26-35 puts average at 31-34 and excellent above 45. The estimate isn't crazy — it lands in the "above average" strip of a forty-year-old dataset. It is just an estimate wearing a table's clothes, and two outlets published two different ones in the same year.
Why did the military stop testing sit-ups?
Because the institutions with the biggest fitness datasets on earth decided it measured the wrong thing. Three branches, three separate calls, inside three years.
The Navy went first. Navy Times reported on 19 November 2020 that the Naval Health Research Center concluded the forearm plank was a better test of core strength and abdominal endurance, because "the repeated spinal flexion movement of the curl-up is not operationally relevant, may aggravate lower-back injuries and does not appropriately challenge the abdominal musculature." Read that again: not relevant to the job, possibly aggravating to backs, and not even a good challenge for the abs. Military.com covered it the day before, noting curl-ups had been in the Navy's test since the 1980s. The plank took effect with the fitness cycle that ran 15 March to 15 September 2021.
The Marines went next. Stars and Stripes, 5 August 2021: the plank was an optional alternative to crunches from 2020, and becomes mandatory in 2023, replacing them outright. The Corps' stated reason lands on the same anatomy the Army's own manual named — "sit-ups and crunches with the feet restrained require significant hip flexor activation," which it said had been "linked to an increased risk of injury, including lower back pain." The plank, it argued, "engages almost twice the abdominal muscles that the crunch does." Max score was set at a 3:45 hold, minimum 1:10.
The Army finished it. Army Times, 23 March 2022: effective 1 April 2022, "the leg tuck is no longer an event — the plank, with a newly lowered standard, is the only core event." When the ACFT became the test of record on 1 October 2022, the two-minute sit-up that had graded soldiers for forty years was simply gone.
Nobody is claiming the sit-up is dangerous. They are claiming it is a bad instrument — it grades your hip flexors and your tolerance for discomfort along with your abs, and it doesn't predict what the test exists to predict.
Are sit-ups bad for your back?
Probably not for you, and the popular version of this claim is a lot wider than the evidence under it.
The number everyone cites comes from Stuart McGill, the spine biomechanist, whose book Low Back Disorders puts the compression of a feet-held sit-up at roughly 3,300 N through the lumbar spine on every repetition — the same territory as the occupational action limit for low-back compression. (McGill states that limit as 3,300 N; the US National Institute for Occupational Safety and Health sets it at 3,400 N, or 770 lb. The argument is the order of magnitude, not a tie.) His underlying lab work is Axler and McGill (1997) in Medicine & Science in Sports & Exercise, which measured spine load across 12 abdominal exercises in nine volunteers and found that partial curl-ups produced the best ratio of abdominal challenge to spinal cost. Real measurement, useful ranking.
Now the hedge, because that number gets used sloppily. The NIOSH action limit is an occupational threshold — repeated lifting across a work shift, in a working population, over years. Not three sets of sit-ups twice a week in a healthy trained lifter. Nine subjects is a mechanics study, not an injury study, and nobody has run the decade-long trial. Plenty of respected coaches argue the flexion panic overshot, and they have a point: your spine flexes when you tie your shoes, and a body never trained to flex under load isn't obviously safer.
So here's the defensible version, narrower than the headline. Sit-ups are a poor test — exactly what three branches concluded from their own data. That is not the claim that sit-ups will hurt you. If your back is already unhappy, pick a different exercise, same as you would for a cranky shoulder. Otherwise the reason to stop doing endless sit-ups isn't fear; it's that they're a weak way to build the muscle. The same trap shows up on how many pushups you should be able to do: a cheap field test mistaken for a training program.
How many sit-ups should I do a day?
Zero, as a daily number. Abs are muscle, and muscle grows from load, not from a rep count you keep inflating.
My own ab work is cable crunches, machine crunches and weighted decline sit-ups — 2 to 3 sets of 6 to 12 reps, twice a week, with the weight going up over time. Same rules as every other muscle I train: four lifting days, 3 sets of 8 to 12 at one to two reps in reserve, add something every session. The part worth noticing is that my version of a sit-up is a weighted one in the 6-12 range, not a timed bodyweight sprint for reps. That isn't a preference. It's the whole argument — if the only variable you can move is how many you did, you have an endurance test, not a training stimulus. Chasing a higher rep count forever is the ab version of curling the same dumbbells for a year and wondering why your arms look identical.
The training side of that is the four exercises that actually show your abs. And the second dial, the one that decides everything: abs show under roughly 15% body fat and are clear at 10-12%, and no rep count changes that number. I went from 285 lbs to about 200 and hold 8-10% year-round, and the visible-abs part was body composition, not volume. Both dials at once is what the Abs Protocol is for — build the muscle, then reveal it.
What should you test instead?
The plank, for the plainest reason available: it is what the Army, the Navy and the Marines replaced the sit-up with. Three institutions sitting on more fitness data than any research group will ever collect, independently swapping a timed flexion test for a timed isometric hold. It is also age-graded and published, which no sit-up standard in current use is. The numbers, the percentiles and the protocol are all in how long you should be able to plank.
But keep a test in its box. Every benchmark on this page — sit-ups, curl-ups, planks — tells you about endurance under a specific protocol, and none of them tells you whether you can see your abs. That is decided by body fat and by how much muscle you built under it. Measure the thing you actually want. If the goal is visible abs, the score that matters isn't on a stopwatch.