What is a good VO2 max by age and sex?

A good VO2 max for your age and sex is one near the average for people who share both, and those averages fall steadily with age. The clearest directly measured set comes from 3,992 healthy adults in Brazil, each walked to exhaustion on a treadmill while a mask measured the oxygen they used [1]. Athletes, smokers, people with obesity and anyone on daily medication were excluded, so this is a healthy reference group rather than a snapshot of the general population.

Averages by age band, in points on that oxygen scale. The first pair in each line is for people who exercise regularly, the second for people who are sedentary, meaning they do not exercise regularly [1]:

• aged 15-24 - 50.6 for men and 38.9 for women; 47.4 and 35.6 if sedentary
• aged 25-34 - 47.4 and 38.1; 41.9 and 34.0 if sedentary
• aged 35-44 - 45.4 and 34.9; 39.0 and 30.0 if sedentary
• aged 45-54 - 40.5 and 31.1; 35.6 and 27.2 if sedentary
• aged 55-64 - 35.3 and 28.6; 30.0 and 23.9 if sedentary
• aged 65-74 - 30.0 and 25.1; 23.1 and 21.2 if sedentary

So a man of 40 who trains most weeks is around average at 45, and around average at 39 if he does not. A woman of 50 sits near her group at 31 if she exercises and 27 if she does not [1].

If you already have a number, read it against your own band rather than the table as a whole. A reading of 42 is below average for a man of 25-34 who trains, and above average for a man of 45-54 who trains [1]. The same 42 sits above every women's group average here, the highest of which is 38.9 [1].

Being a point or two off your band's average means little, because people inside each group were spread widely, by about 3 to 8 points either side. Roughly two-thirds of sedentary men aged 45-54 landed between about 28 and 43 [1].

Women averaged below men in every band, which is why a mixed-sex chart misleads. The gap was close to 12 points at 15-24 and under 5 by 65-74 [1].

Other reference sets land differently. In 18,189 Brazilian treadmill tests, the middle score for men fell from 44.7 at 20-29 to 28.4 at 60-69, and for women from 36.3 to 22.3 [2]. In 4,199 Chinese treadmill tests, men averaged 42.6 at 20-29 and 35.9 at 50-59, women 37.1 and 30.3 [4]. Among about 12,000 British adults screened at private clinics, median age 48, the middle half of the 9,204 men scored between 30.5 and 44.7 and the middle half of the 2,687 women between 30.1 and 44.8 [5]. Those women sit far closer to the men than either Brazilian set would predict.

Whichever chart you use, the direction is the same, and that is why the number gets tracked. An overview of 199 cohort studies covering 20.9 million observations found the fittest people were about half as likely to die during follow-up as the least fit [11]. Each extra MET, the conventional unit for the oxygen the body burns sitting still and worth about 3.5 points, went with roughly 11 to 17% lower risk of dying from any cause [11]. A separate pooling of 34 cohort studies put that per-MET figure at about 12% [12].

How far your own number can move depends on where it starts. Seven weeks of interval training raised the peak reading by 11.4% in one group of trained recreational cyclists who started around 52 [15], while four weeks of it left the figure unchanged in 17 well-trained runners [16].

Oxygen delivery, not lung size, sets the ceiling. A review of the underlying biology argues that training-driven rises in VO2 max come mostly from a larger volume of red blood cells, the cells that carry oxygen around the body, and a bigger stroke volume, the amount of blood the heart pushes out with each beat [13].

The same review reports that every previously untrained person in the studies it covers improved, provided the training passed some threshold of volume or intensity, and that how much each improved appears to track those same circulatory factors [13].

A high number does not by itself mean high endurance. Among competitive athletes, VO2 max works alongside the lactate threshold, the effort level above which lactate builds up in the blood faster than the body clears it, and efficiency, the oxygen cost of holding a given running speed or cycling power [14].

Two athletes with identical VO2 max values can perform very differently because of those other two factors [14].

Every table here describes the population it came from, and those populations disagree with each other. Age-related values in healthy Koreans differed from Western ones, and the authors concluded that fitness should be judged against national or ethnic standards rather than one global chart [6]. In the 18,189-test Brazilian set, Norwegian men and women scored above both the American and Brazilian groups in every age band [2].

The reference points move over time as well. The 2022 American standards came out 1.5 to 4.6 points lower than the 2015 version they replaced [3]. In 354,277 Swedish working adults tested between 1995 and 2017, average fitness dropped by 4.2 points and the share below the study's low-fitness cut-off of 32 rose from 27% to 46%, about a third of that tracking rising body weight [8]. Matching today's average is a lower bar than matching the average of 1995.

The machine changes the number too. In 15,045 tests from an American registry of people with heart disease, treadmill results averaged 21% higher than cycle results [9]. A cycle score and a treadmill score are not interchangeable.

Only some of these figures were measured directly. The British numbers were estimated from a submaximal treadmill test and the Swedish ones from a submaximal cycle test, meaning both were worked out from how the body responded at moderate effort rather than read off a maximal test [5][8].

If your number came from a wrist device, treat it as a rough guide. Against laboratory measurement in 40 healthy adults, Apple Watch Series 10 placed only 5 people in their correct fitness band, with 12 more one band away [10]. It read about 6 points low on average, and individual readings scattered widely around the true value [10].

The Brazilian table excluded athletes, smokers, people with obesity and anyone taking daily medication [1]. The Nigerian equations come from 124 adults tested in the 1980s, and none could be fitted for active women because none over 29 volunteered [7].

The mortality figures describe associations measured across whole populations. They are not evidence that raising one person's number by 3.5 points cuts that person's own risk by 11 to 17% [11][12].

None of these papers report how much the same person's score moves between two tests, so they cannot say how large a change has to be before it beats ordinary measurement noise.

Who these numbers do not describe:

• anyone under 15 or over 74, the outer limits of the Brazilian age bands [1]
• competitive endurance athletes, whose results depend on lactate threshold and efficiency as much as on VO2 max [14]
• people with heart or lung conditions. In 16 adults with a Fontan circulation, a surgically rerouted heart with one working pumping chamber, the peak oxygen figure averaged 66% of what would be expected for their age and size [17]

Take a 45-year-old man whose watch reports 42.

Against the Brazilian table his band is 45-54, where men who train averaged 40.5 and men who do not averaged 35.6, so he reads as above average either way [1]. Against the British screening data, where the middle half of 9,204 men ran from 30.5 to 44.7, he sits inside the middle half rather than above it [5].

The watch is the looser part of that. Apple Watch Series 10 read about 6 points below the laboratory measurement on average in 40 healthy adults, which would put his measured figure nearer 48 [10]. Agreement on individual readings was poor enough that his true value could sit anywhere from the mid-30s to around 60 [10].

A repeat reading of 44 next year would be a gain of well under one MET, about 3.5 points. The mortality figures attached to 1 MET were measured across whole populations, not inside one person over one year [11][12].