VO2max by age and sex: what the reference data shows

On lab-measured US data, the median VO2max is 48.0 ml/kg/min for men in their 20s and 37.6 for women, falling to 24.4 and 18.3 by their 70s.¹ Those numbers come from the FRIEND registry, a US database of real treadmill tests with a gas analyser, and they're about the fairest yardstick most of us have. Most "VO2max by age" charts online don't say where their numbers came from, and the published tables don't agree with each other either.

So below is the FRIEND table itself, why the other tables disagree with it, and where the cut points hybridAF grades against actually sit. That last part is less flattering to us than you might expect.

The FRIEND percentiles#

FRIEND's 2015 report pooled 7,783 maximal treadmill tests from eight US labs, all adults aged 20 to 79 without known cardiovascular disease, and every test had to reach a respiratory exchange ratio of at least 1.0, a standard check that the effort was close to maximal.¹ That's 4,611 men and 3,172 women. VO2max here is measured, not predicted: oxygen in, oxygen out, through a mask.

These are the treadmill percentiles from Table 3 of that paper, in ml/kg/min.

The paper also reports the 5th, 10th and 95th percentiles. If you're a man in your 20s, 66.3 puts you at the 95th. For women in their 20s it's 56.0.

Read the age columns carefully. These are different people of different ages tested at one point in time, not one group followed for 50 years. The roughly 10% drop per decade is a cross-section, and it says nothing about how fast your own number has to fall if you keep training.

Two more things fall out of the same paper. The gap between men and women shrinks with age, from 10 ml/kg/min in the 20s to 7.5 in the 70s, because the men's numbers fall faster: about 0.44 ml/kg/min a year against 0.38 for women.¹ And the youngest and oldest bands are the thinnest. Only 513 men and 410 women were in their 20s, and 137 men and 98 women in their 70s, so the extreme percentiles in those rows rest on fewer people.

The 2022 update moved the line down#

In 2022 the FRIEND group published updated standards from 22,379 tests across 34 labs, now including cycle tests and people up to 89.² The treadmill standards came out 1.5 to 4.6 ml/kg/min lower than the 2015 ones, and the average decline was 4.0 ml/kg/min per decade on the treadmill. Cycle tests came out lower than treadmill tests, which matters if your lab number came off a bike.

The table above is the 2015 one because it's open access and every cell could be checked against the paper. The 2022 tables weren't accessible for this article, so none of its percentiles appear here. The direction is what matters: on the newer US data, the same VO2max lands at a slightly higher percentile than it does above.

Why other charts disagree#

Reference tables describe the people who were tested, and those people vary a lot by country and by who walked into the lab.

FRIEND's own paper puts its means next to a Norwegian sample (Loe et al.), and the gap is large: men in their 20s averaged 47.6 in FRIEND and 54.4 in the Norwegian data, women 37.6 and 43.0.¹ The FRIEND authors' conclusion is blunt: normative values are region and country specific. Their own sample isn't a random draw either. People were tested for clinical assessment, fitness screening or research, at labs in eight states.

The same paper lists Cooper Clinic values, which many older charts copy. Those were predicted from treadmill test time, not measured with gas analysis, and they're flatter across age: 43.9 at the median for men in their 20s, 30.9 in their 70s.¹

There's also a third kind of chart: the ones built from a formula. Norway's HUNT study measured 4,637 adults and built a model that predicts VO2peak from age, waist, activity and resting heart rate, with a typical error of 5.70 ml/kg/min in men and 5.14 in women.³ That model is behind NTNU's free online fitness calculator.⁴ It's a good guess for someone with no training data. It isn't a measurement.

What about runners and HYROX athletes#

FRIEND is a general-population table. Nobody in it was selected for training, so if you run or race seriously, the 75th and 90th percentile columns are the comparison that tells you something. The median mostly tells you you're not sedentary.

There's very little lab data on hybrid athletes specifically. The one HYROX lab study so far measured 11 recreational HYROX athletes before a simulated race: their median VO2max was 51 ml/kg/min, ranging from 46.6 to 72, with a median age of 33.⁵ That's a tiny sample, but it gives you a feel for the field. On FRIEND, 51 for a man in his 30s is around the 80th percentile.

Where hybridAF's cut points sit#

The hybridAF Engine Index grades VO2max against five cut points per age band, mapped onto a 20, 40, 60, 80, 95 points ladder (untrained, novice, intermediate, advanced, elite). The cut points are a design choice of ours, not a published table, and they're under review. Here's how they line up against FRIEND 2015.

FRIEND median vs hybridAF intermediate

age band midpoint	FRIEND men, 50th (ml/kg/min)	hybridAF men, intermediate (ml/kg/min)	FRIEND women, 50th (ml/kg/min)	hybridAF women, intermediate (ml/kg/min)
25	48	48	37.6	41
35	42.4	45	30.2	39
45	37.8	42	26.7	36
55	32.6	39	23.4	33
65	28.2	35	20	30

FRIEND 2015 treadmill 50th percentile (Kaminsky 2015, Table 3) against the hybridAF intermediate cut point (src/lib/engine-grading.ts). hybridAF's 60+ band is plotted at 65.

For men under 30 the match is exact: our intermediate cut point is 48, FRIEND's median is 48.0. After that the two lines split. Our cut points step down about 3 ml/kg/min per decade, and FRIEND's median drops 4 to 6. By the 50s our "intermediate" for men (39) sits near FRIEND's 73rd percentile.

For women the gap is bigger and starts earlier. Here's every women's cut point placed on FRIEND 2015, approximately, by interpolating between the published percentiles.

Put plainly: a 45-year-old woman at the FRIEND median, 26.7, scores about 28 on our ladder, between untrained and novice. Our women's cut points sit much closer to the Norwegian means (women in their 40s averaged 38.4 there, against our intermediate of 36¹), and the men's fall between the US and Norwegian figures. But they're ours. Nobody published them, and it's better to know the yardstick is strict up front than to find out by comparing tables.

What that means for you: if the app calls you novice and FRIEND puts you near the middle, both can be true. They're answering different questions. FRIEND tells you where you sit among Americans who got a lab test. Our ladder is the one the strength side of the score is graded on too, so the two halves of a hybrid number mean the same thing. You can see both sets of standards side by side on the hybrid athlete standards page.

A lab number and a watch number aren't the same thing#

Everything above is measured VO2max. Your watch's number is an estimate from heart rate and pace, and a pooled analysis of 14 validation studies put the 95% limits of agreement for exercise-based wearables at roughly 10 ml/kg/min either side of the lab value.⁶ On the FRIEND table, 10 ml/kg/min takes a man in his 30s from the 25th percentile (35.9) to about the 63rd. One reading can't place you that precisely.

So read your number as a range, then look the whole range up. The Apple Watch and Garmin accuracy breakdown goes through what the validation studies found for each.

A worked example: my own number#

My current estimate, read on 22 October 2026, is 47 ml/kg/min with a range of plus or minus 5, from the heart rate and pace of 10 recent steady runs. I'm a man in the 20 to 29 band.

On FRIEND 2015 that point estimate sits around the 47th percentile, just under the median of 48.0. The range is the honest part: 42 lands around the 31st percentile and 52 around the 64th. So the most I can say is that I'm somewhere in the middle third of US men my age who've had a lab test. On the hybridAF ladder the same 47 grades as an Engine Index of 57, a touch under intermediate, which matches FRIEND for my age band because that's the one band where our line and theirs meet.

What the number is and isn't#

VO2max is the ceiling of your aerobic engine. It's limited mostly by how much oxygen your heart and blood can deliver, and it sets the upper limit on endurance performance, but it doesn't decide a race on its own: running economy and the share of your VO2max you can hold matter too, and speed at lactate threshold predicts distance running better than VO2max does.⁷ Two runners with the same VO2max won't necessarily run the same 10K.

So use the chart for what it's good at: a rough sense of where your engine sits for your age and sex, and a baseline you can retest the same way later. If you want to move it, raising VO2max without losing strength covers what the training studies found. If you race HYROX, the lactate threshold vs VO2max piece is probably the more useful read. The rest of the engine articles live on the engine hub.