How to improve VO2 max: the training that works, and how long it takes

Short answer

VO2 max responds to intensity, not to volume alone. The protocol with the best evidence behind it is four intervals of four minutes at 90–95% of maximum heart rate, with three minutes of easy recovery between — the Norwegian 4×4. In the original trial it raised VO2 max by around 7–10% over eight weeks.

If you have been running steadily for months and the number has not moved, that is not your fault and not a sign you lack discipline. Steady easy work builds a great deal, and it does not, on its own, push this particular ceiling.

Note: this article is about training and is not medical advice. High-intensity intervals are not appropriate for everyone. If you have a heart condition, uncontrolled blood pressure, or have been inactive for a long time, speak to a clinician before starting. If exertion makes you worse a day or two later, this article does not apply to you — see the warning below.

What actually raises VO2 max

One mechanism does most of the work, and knowing it explains why the protocols look the way they do.

VO2 max is limited mostly by how much blood your heart can pump per minute — and specifically by stroke volume, the amount ejected per beat. Training raises it by making the heart spend time near that maximum, which forces adaptation.

The catch is that you only reach near-maximum stroke volume when you are working very hard, and you cannot stay there long. So the training problem is: how do you accumulate the most minutes near your ceiling before you have to stop?

The answer is intervals. Hard enough to get there, long enough to stay, repeated with just enough recovery to go again.

This is why easy volume alone plateaus. Zone 2 work is genuinely valuable — it builds mitochondrial density, fat oxidation, capillary networks and the base that makes hard sessions possible, and most of the cardiovascular health benefit accumulates there. But it does not spend time near maximum stroke volume, so the ceiling itself moves slowly.

The protocol with the best evidence

The Norwegian 4×4 is the most studied single session in this field, developed by Jan Helgerud and Jan Hoff at NTNU and studied extensively by the cardiac exercise research group at the same institution.

The structure:

About 38 minutes in total.

What it produced: in Helgerud's trial, roughly 7 to 10% improvement in VO2 max over eight weeks at three sessions a week. The 4×4 arm outperformed both steady-state running and shorter intervals in the same study.

Two practical notes that matter more than the exact numbers.

Four minutes is not arbitrary. Shorter intervals do not give the heart long enough near maximum stroke volume; much longer ones force the intensity down. Four minutes is where those two constraints meet.

Three sessions a week was the trial, two is usually the realistic version. Most of the gain shows up at two sessions, with far better odds of still doing it in month three. A protocol you abandon in week four produces nothing.

How hard is 90–95% of maximum?

Hard enough that people routinely underestimate it, which is the most common reason the session does not work.

By feel: you should be able to speak two or three words, not a sentence. By the last minute of each interval you should genuinely want it to stop.

By heart rate: if you are using a monitor, expect your heart rate to climb through the interval rather than sit at target from the start. The first interval often looks too low on the watch even when the effort was right — cardiac output takes time to rise.

The estimated-maximum problem. Most people are working from a formula, and max heart rate by age formulas carry a wide margin of error for individuals. If the numbers feel wrong for you, trust the effort over the calculation.

And a useful check: if you could have done a fifth interval comfortably, the first four were too easy. If you could not finish the third, they were too hard.

If you are on a beta-blocker, percentage-of-maximum targets stop meaning anything, because the drug caps the response the target is calculated from. Effort becomes the only usable guide, and this is worth raising with whoever prescribed it before starting a block of intervals.

How long does it take?

First changes: two to four weeks. Usually felt rather than measured — the same effort at a lower heart rate, recovering faster between intervals.

Measurable improvement: six to twelve weeks. The eight-week window in the research is about right for a meaningful shift.

The fastest gains go to the least fit. Someone starting from a low base can improve substantially; a trained athlete may gain a few percent and then very little. This is worth knowing in both directions — encouraging if you are starting out, and a reason not to keep chasing the number if you are already trained.

And it goes away. Detraining is faster than training. Several weeks off undoes a meaningful share of what took months to build, which is an argument for a maintenance session rather than for training in blocks and stopping.

What a week actually looks like

The mistake is doing only the hard part. The research protocols sat on top of a base, and so should yours.

Two hard sessions a week, maximum. More is not better here, and three is the ceiling even in the trials.

Everything else easy — and genuinely easy. The most common error in amateur training is the moderate middle: sessions too hard to be recovery and too easy to be a stimulus. Easy should feel almost boringly conversational.

Forty-eight hours between hard sessions. The adaptation happens in the recovery, not the session.

Strength work twice a week does not raise VO2 max directly but supports everything around it, and matters increasingly with age.

One deliberately lighter week roughly every fourth. Accumulated fatigue suppresses performance and makes the hard sessions worse.

Sleep is part of the plan, not adjacent to it. Adaptation to a hard session happens overnight, and a short night after intervals buys you the cost without the benefit. If something has to give in a busy week, drop the second hard session rather than an hour of sleep.

Food around the session matters more than supplements. Carbohydrate available before a hard interval session is the difference between four good intervals and two good ones followed by two survival efforts. This is not a nutrition article, but under-fuelling is a common and invisible reason a block underperforms.

The intervals can be run, cycled, rowed, or done on a hill or a bike. The heart does not know what you are doing — it responds to the demand. Choose the mode your joints tolerate and your life accommodates.

The alternatives to 4×4, and when they suit better

The 4×4 has the best evidence, not a monopoly. Several other structures work, and the choice is mostly about what you can tolerate.

Shorter intervals — 30/15s or one-minute repeats. Thirty seconds hard, fifteen easy, repeated in blocks; or a minute on, a minute off. These accumulate time near maximum too, and many people find them psychologically easier because no single effort lasts long. The trade-off is more total repetitions and a heart rate that dips between them.

Longer intervals — 5 to 8 minutes. Closer to threshold work than pure VO2 max stimulus, and excellent for the other half of performance, which is how much of your ceiling you can hold. Lower intensity, higher sustainability.

Hill repeats. Effective, joint-friendlier than flat sprinting for some, and self-limiting — the hill enforces the intensity so you cannot drift.

Fartlek, or unstructured hard efforts. Less precise and much better than nothing. For people who will not run to a stopwatch, this is the version that actually happens.

Two things decide between them, and neither is which is theoretically optimal.

Can you do it on the equipment and terrain you have? Four uninterrupted minutes needs somewhere to run or ride without stopping at junctions.

Will you still be doing it in eight weeks? The protocol you complete twenty times beats the better protocol you complete four times. Every improvement figure quoted in this field assumes adherence, and adherence is where most plans actually fail.

How to fit this in if you are short on time

The honest version, because "38 minutes, three times a week" does not survive contact with most lives.

The minimum that works is two sessions a week. Below that, gains become slow and unreliable.

A single session can be cut to three intervals without losing much. Three by four minutes plus a shortened warm-up is around 28 minutes and still a real stimulus. Cutting the warm-up entirely is the wrong economy — going hard cold is both less effective and more likely to hurt.

Commuting counts if it is hard enough. A cycle commute with two genuinely hard four-minute segments is a session. It does not need to be labelled training.

Stairs work. Four minutes of stairs is unpleasant, requires no equipment and reliably gets heart rate where it needs to be.

And if you miss a week, resume rather than restart. The base does not vanish in seven days, and treating a missed week as a failure is how eight-week plans become three-week plans.

Why your watch number may not move

You can be improving while the estimate sits still, and this frustrates people into abandoning a plan that is working.

The estimate is soft. A watch infers VO2 max from heart rate and pace without measuring oxygen, so it carries noise that can easily swamp a 5% change. What a good VO2 max is covers where those numbers come from and why they differ from lab figures.

It needs the right kind of session. Most models want steady outdoor running or walking with good GPS. Intervals, treadmills, cycling and rowing often produce no update at all — so the training that raises VO2 max is frequently the training your watch cannot read.

It lags. Estimates are smoothed over weeks. A real improvement shows up later than it happens.

Weight moves it without fitness changing, because the measure is per kilogram.

The better short-term evidence is elsewhere: resting heart rate drifting down, heart rate recovery after intervals getting faster, and the same pace costing fewer beats. Those respond in weeks and they are harder to fake.

The mistakes that waste the eight weeks

Most failed blocks fail the same handful of ways.

Going too hard on the easy days. The single most common error. Easy sessions creep up to moderate, which is too hard to recover from and too easy to stimulate anything, and then the hard sessions are performed tired and are no longer hard. The fix is uncomfortable: make the easy days slower than feels dignified.

Going too easy on the hard days. The mirror image. Four minutes at a pace you could hold for twenty is not a VO2 max interval, whatever the watch reports afterwards.

Cutting the recovery between intervals. The three minutes are there so the next interval can be done at full intensity. Shortening them makes the session harder and less effective at the same time, which is the worst possible trade.

Adding sessions instead of adding quality. Four hard sessions a week does not double the result; it produces accumulated fatigue and a plateau. Two done properly beats four done tired.

Changing everything at once. New shoes, new diet, new schedule and intervals in the same fortnight means you will not know what worked or what hurt.

Judging it by the watch estimate at week three. Too early, wrong instrument, and the most common reason people abandon a plan that was working.

And skipping the base. Intervals on no aerobic foundation is how people acquire an injury in week two and conclude that intervals are dangerous.

When not to do this

This matters more than the protocol, and it is usually a footnote elsewhere.

If exertion makes you worse a day or two later, stop reading this as advice. In ME/CFS and long COVID with post-exertional malaise, high-intensity intervals cause harm, and raising VO2 max is the wrong goal. Pacing within an energy envelope is the approach that applies. This is not a matter of starting more gently — the direction is wrong.

If you have a heart condition, uncontrolled blood pressure, or unexplained chest pain or breathlessness, this needs a conversation with a clinician first, not a modified plan.

If you have been sedentary for years, build several weeks of easy aerobic work before adding intervals. The base is not optional and skipping it is how people get injured in week two.

If you are ill, or badly short of sleep, skip the hard session. A hard interval session on top of an infection is a bad trade.

And if you are already well trained, accept that the returns are small. Further gains in performance come from raising the fraction of your ceiling you can sustain, not the ceiling itself.

How to bring this up with your doctor

Ask directly whether high-intensity intervals are safe for you, and say what you mean by that — four minutes at near-maximum effort, twice a week. "Should I exercise more" gets a different answer than the specific question.

Mention medication, particularly beta-blockers, which cap heart rate and make percentage-of-maximum targets meaningless. You will need effort-based targets instead.

Raise any of these before starting: chest pain or pressure on exertion, breathlessness out of proportion, dizziness or fainting with effort, palpitations, a family history of sudden cardiac death under fifty.

If you have been told to improve your fitness and given no detail, ask for specifics — intensity, frequency, what to do if symptoms appear, and whether an exercise test is warranted first.

And stop and seek help for chest pain, severe breathlessness, fainting or an irregular heartbeat during exercise. That advice is dull and it is the reason this section exists.

How Welltory helps — and what it cannot do

The limit first. Welltory does not measure VO2 max, does not prescribe training and does not clear you for exercise. It is a general wellness product with no regulatory clearance and it diagnoses nothing.

What it does is answer the question a training block actually raises, which is not "has my VO2 max gone up" but "is my body absorbing this or accumulating damage?" Those are different questions and the second one is answerable sooner.

Resting heart rate, heart rate variability and overnight recovery respond to training load within days to weeks — long before any VO2 max estimate moves. Read against your own months of history, a downward drift in HRV alongside a rising resting heart rate across a fortnight is the classic picture of load exceeding recovery, and it is the most useful thing a wearable can tell someone doing intervals twice a week.

Two honest caveats. These signals are non-specific — illness, alcohol, heat and poor sleep move them the same way overtraining does, so read the trend and not the morning. And a low reading is information, not an instruction: sometimes the right response is a rest day, sometimes it is that you had a glass of wine.

A record of how the training is landing, not a coach.

How we made it

Made with AI tools, then edited and fact-checked by the Welltory team. See our Editorial & AI policy.

Data analysis by Jane Smorodnikova, co-founder of Welltory and the person who built the methodology behind how we read physiological data.

Written by Tatsiana Yashyna.

Discounts for blog readers: up to 36% off

See what affects your energy, stress, sleep, and daily state with Welltory

Explore your data