How to Improve Your VO2 Max, and How Much Is Realistic

VO2 max is how much oxygen your body can take in and use in a minute of all-out work, divided by what you weigh. It is the best-studied measure of fitness there is, and in a cohort of 122,007 people followed for a median of 8.4 years, being in the bottom quarter of it carried a larger adjusted death rate than smoking, diabetes or established coronary artery disease did in the same model.

So the question of how to raise it is worth answering properly. The short answer is that you have to be uncomfortable, and that how much you gain is partly out of your hands.

The trial that settled the intensity question#

In 2007 a Norwegian group took forty moderately trained men and split them into four training methods, each done three times a week for eight weeks. The important design choice was that all four groups were matched for total oxygen consumed across the study, so nobody could win by simply doing more work.

The two easier methods, done for the same total work, did not move the number enough to be detectable. The two interval methods did, and the four-by-four did it best. Stroke volume, the amount of blood the heart pushes out per beat, went up about 10 percent in the interval groups and not at all in the others.

That study is small, it was young men, and eight weeks is short. But it is the cleanest work-matched comparison available, and every larger meta-analysis since has pointed the same way, with a smaller effect. Pooling 28 studies and 723 participants found intervals better than steady endurance work by about 1.2 ml/kg/min, and the advantage was largest in the people who started least fit.

The four-by-four, exactly#

Warm up for ten minutes easy. Then four times: four minutes at an effort where you could not hold a conversation, followed by three minutes easy enough to recover. Cool down for three minutes. Three times a week.

The hard four minutes should feel like something you could sustain for perhaps six or seven minutes if you had to, not a sprint. The first interval usually feels too easy and the last one usually feels too hard, which is roughly correct.

It works on a bike, a treadmill, a rowing machine, a hill, or a flight of stairs. The equipment matters far less than the honesty of the effort.

What you can actually expect to gain#

Here the literature does something unusual: it tells you clearly that the answer varies enormously between people.

The HERITAGE Family Study put 481 sedentary adults from 98 families through twenty weeks of identical supervised cycling. The average gain was meaningful. The individual range ran from essentially zero to very large, and about 47 percent of the difference in trainability was inherited.

That finding deserves to be said out loud, because most training advice implies a guarantee. Some people on a perfect program gain little. If that is you, it is not a character failure, and it does not mean the training is doing nothing for your blood pressure, your blood sugar, your sleep or your mood.

For rough planning, a sedentary adult can expect 15 to 20 percent over a few months, which for most people is six or seven points and enough to move up a percentile band or two. That is about what two decades of average age-related decline costs, so it is a meaningful thing to buy back.

Where zone 2 fits, honestly#

Easy aerobic work has become the dominant message in longevity circles, and the strongest version of the claim, that it is specifically the best intensity for building mitochondria and raising fitness, is not supported.

A 2025 review looking at exactly this question concluded that current evidence does not support easy aerobic work as the optimal intensity for improving mitochondrial capacity, and that several studies found no effect at all. Pooled data suggest that work below about 60 percent of your maximum may be below the threshold for building mitochondrial content in the first place.

The defensible version is different and still worth knowing. Mitochondrial content tracks accumulated training volume, and easy work is how you accumulate volume without needing days to recover. That is a real rationale for doing most of your training easy. It is just not the same claim as easy work being uniquely good for you.

The practical shape for someone with three or four hours a week: keep most of it comfortable, and put one or two properly hard sessions in. The eighty-twenty split that endurance athletes use was worked out by people training fifteen to twenty-five hours a week, where twenty percent hard is still three to five hours.

Scaled down to 150 minutes a week, twenty percent is half an hour, and the subgroup data suggest people with low volume and low fitness get more from a larger share of intensity, not less.

Walking, and where it stops#

For someone sedentary, walking works. Across 32 randomised trials of previously sedentary adults, walking programs raised VO2 max by about 3 ml/kg/min, roughly a 10 percent gain, alongside about 3.6 mmHg off systolic blood pressure and a little over a kilogram of body weight.

The reason it works is the reason it stops working. In a deconditioned person, ordinary walking is a high relative effort. As fitness rises, the same walk becomes easy, and an easy walk is no longer a stimulus. The Norwegian trial's easy-pace group, who were already moderately fit, gained 1.8 percent and it did not reach significance.

So walking is an excellent place to start and a poor place to stay. When it stops feeling like anything, add hills, add pace for part of it, add a loaded pack, or add intervals.

Be careful with prescribed heart rates#

Most training plans hand you a heart rate band derived from your age. The arithmetic behind it is shakier than it looks.

The two best-validated equations, 208 minus 0.7 times your age and 211 minus 0.64 times your age, disagree by about 7 beats for a 45 year old before any individual variation. The second carries a standard error of 10.8 beats per minute, and the researchers who derived it note that fitness, sex, activity level and body mass do not help you predict your own maximum any better.

A typical training band is ten to seventeen beats wide. The prediction error is larger than the band. For a meaningful number of people, a heart rate zone calculated from age is simply in the wrong place.

Use effort instead. The intensity where speech stops being comfortable is a better anchor than a number, and it is free.

What this is not#

None of this is a promise about your lifespan. The fitness and mortality association comes from watching people, not from randomising them, and it is measured in people referred for stress testing rather than in the general population. Higher fitness travels with a great many other things that are good for you.

What is reasonable to say is that fitness is one of the most modifiable things on the list, that the gap between the bottom and the top of the distribution is enormous, and that the training which closes it is well described and takes about half an hour, three times a week.

Find out where you are starting from with the VO2 max estimate, and see the other five tests that read the same underlying reserve.