If you are only going to measure one thing about your fitness, measure this one. And if you are going to train one thing deliberately, this is the strongest candidate, with a caveat about what raising the number does and does not prove, which I will get to.
The good news for a beginner: the best-studied protocol in this entire field is four intervals of four minutes. You can learn it in a minute. Earning the right to do it takes about twelve weeks, and that part matters more.
⚠ Before anything here. This page describes deliberately hard exercise at 90–95% of maximum heart rate. If you have a diagnosed heart condition, uncontrolled blood pressure, chest pain on exertion, or you are over 40 and have been sedentary for years, get cleared by a doctor before doing interval work. This is not boilerplate. It is the one page in this series where the intensity itself carries risk, and “I felt fine walking” is not a clearance.
Where to start, depending on who you are
| If you are | Go to |
|---|---|
| New to running entirely | Not here yet. Start with Zone 2, then come back |
| Running easily for 8+ weeks, no niggles | The twelve-week ramp: you are ready to begin building toward it |
| Wanting to know what VO2 max even is | What it is, first |
| Already doing intervals randomly | The 4×4: you probably have the intensity wrong in one direction or the other |
| Over 50 or long detrained | The twelve-week ramp, and take sixteen |
What VO2 max is
The maximum amount of oxygen your body can take in and use during hard exercise, expressed per kilogram of bodyweight per minute. It is the ceiling on your aerobic engine: how much work you can do before you are simply out of capacity.
It is worth measuring because it is one of the few fitness numbers that is objective, comparable, and trainable. You cannot talk yourself into a better VO2 max.
⚠ A limit worth stating early. VO2 max is strongly associated with living longer, and that association is one of the more striking in the literature. But it is a surrogate. Most of the training research measures whether VO2 max went up, not whether anyone lived longer. The single randomised trial that used this exact protocol and tracked deaths is Generation 100, and it came back largely null on mortality. The reasons are set out on the training plan and they are more interesting than damning. But raise your VO2 max because it makes your life easier and your fitness measurable, not because a trial proved it buys you years.
What the research says, and the one protocol worth knowing
The best study: Helgerud et al., 2007. Researched · randomised, work-matched comparison of four training approaches · n=40 · 8 weeks
Why it wins. It is a randomised comparison of four different training approaches at matched total workload, which is the crucial design feature, because it isolates how you train from how much. Forty moderately-trained men, three sessions a week, eight weeks, with VO2 max measured directly. Most training studies compare a training group to a group doing nothing, which only proves that training beats not training. This one asks the harder question.
What it found. The 4×4 produced roughly a 7% rise in VO2 max in eight weeks, and beat both moderate continuous training and training at lactate threshold at the same total workload. In other words: for this outcome, the intervals were not just more time-efficient. They were more effective for the same work done.
Researched · replicated in obese and coronary-artery-disease populations ### ⚠ But 4×4 is the best-studied protocol, not a magic formula
A balance correction, added 31 August 2026. The first version of this page implied the 4×4 specifically was the answer. That overstates it.
What the evidence actually supports is the principle underneath it: what drives VO2 max is accumulating enough time near your maximum, using long intervals of two minutes or more. Other protocols that achieve the same time-near-max (6×3 minutes, 8×2, 4×8) produce broadly similar adaptations.
What does hold up:
- Long intervals beat short ones for this outcome. In a direct comparison, 4×4 raised VO2 max about 6.5% versus about 3.3% for sprint intervals of 8×20 seconds. When pooled analyses looked at what predicted benefit, it was long-interval (≥2 min), higher-volume (≥15 min of work), sustained over 4–12 weeks. Short, token intervals did not deliver.
- 4×4 has the largest evidence base, which is a real advantage. It means the result is less likely to be a fluke, and the protocol is known to be tolerable.
So use the 4×4 because it is well-studied and easy to follow, not because four minutes is a magic number. If 4×8 or 6×3 suits you better, the evidence does not say you are doing it wrong.
Researched · replicated in obese and coronary-artery-disease populations It replicates across populations, which matters more than any single result. In participants with obesity, six weeks of the same protocol produced around a 10% increase. In patients with coronary artery disease, ten weeks of it beat work-matched moderate training. It is the most-studied HIIT protocol for improving VO2 max, and consistency across different groups is what separates a real effect from a fluke.
⚠ The limits. Forty men in the original, so small and not diverse. Eight weeks is short. And the endpoint throughout is VO2 max, a surrogate. This is good evidence that the 4×4 raises a number that matters; it is not evidence about your lifespan.
How much it will work for you, the part I left out
This is the most important omission from the first version of this page, and it is the sort of thing that makes people quit.
That “~7% in eight weeks” is a group average, and it hides enormous individual variation. In the HERITAGE Family Study, which put families through an identical supervised training programme and measured the response:
- The average gain was meaningful, but roughly 10–20% of people showed limited trainability.
- About 7% gained almost nothing (under 100 ml/min), while about 8% gained more than seven times that.
- Response ran in families. There was about 2.5× more variance between families than within them, and the heritability of trainability itself was estimated around 47%.
Read that plainly: how much your VO2 max improves from a standard programme is substantially inherited, and a real minority of people barely respond to the textbook protocol.
Why this matters more than it might seem. If you run twelve weeks of intervals and your number barely moves, the overwhelmingly common conclusion is “I did it wrong” or “I am lazy.” Often neither is true. You may be at the low-response end of a distribution nobody told you existed.
What to do about it:
- Judge by more than one measure. Resting heart rate, how the same run feels, what you can do on a hill, recovery between sessions. VO2 max is one number, not the verdict.
- Low responders to one stimulus are not necessarily low responders to another. If intervals do nothing over a proper block, try more volume instead, and vice versa.
- The health benefits of training are not conditional on your VO2 max moving. Strength, balance, metabolic health and mood do not check your trainability genes first.
⚠ This cuts both ways and I want to be even-handed about it. Genetics explaining part of the response is not a reason to skip training. The non-responders in these studies were still training, and being at the low end of a training-response distribution is not the same as getting nothing.
The 4×4, exactly
Once you have earned it:
| Warm up | 10 minutes easy, building gradually. Do not skip this; going hard from cold is how people get hurt |
| Work | 4 minutes at 90–95% of max heart rate. Hard. You can say a word, not a sentence |
| Recovery | 3 minutes easy at roughly 60–70%. Keep moving, do not stop dead |
| Repeat | 4 times total |
| Cool down | 5–10 minutes easy |
| Total | About 40 minutes |
| Frequency | Once or twice a week. Never on consecutive days |
How hard is 90–95% actually? By the end of each four minutes you should be working hard and glad it is ending, but able to complete all four intervals at the same effort. If interval four falls apart, interval one was too hard. That is the single most common error, and pacing it out is a skill that takes a few sessions.
Do not use a formula for max heart rate. 220-minus-age is off by ±10–12 beats for an individual. Use effort, or use the highest heart rate you have actually seen on a hard effort.
It works on anything: running, cycling, rowing, an incline treadmill walk, stairs. Cycling and rowing are gentler on joints and are the better choice if you are heavier or have knee history.
The twelve-week ramp: earning the right
Do not start here on week one. Interval training on an untrained body is the fastest route to injury in this entire series.
Prerequisite: roughly eight weeks of consistent easy running, able to hold 30 continuous minutes at conversation pace, no niggles. If that is not you, go to Zone 2.
| Weeks | What you add | Keep doing |
|---|---|---|
| 1–4 | Strides. After one easy run a week, add 4–6 × 20 seconds at a brisk-but-controlled pace, walking back between. Teaches your legs speed with almost no injury risk | 2–3 easy sessions |
| 5–8 | Long intervals, moderate. One session a week: 4 × 3 minutes at “comfortably hard” (around 80–85%) with 3 minutes easy between | 2 easy sessions |
| 9–10 | Bridge. Same session, but 4 × 4 minutes at 85–90%. This is the 4×4’s shape at slightly lower intensity | 2 easy sessions |
| 11–12 | The real thing. 4 × 4 at 90–95%, once a week | 2 easy sessions |
| Then | Hold at once a week. Consider a second session only after a month, and only if recovery is good | 2 easy sessions |
Rules:
- Every fourth week, go easy. Drop the hard session entirely. This is where adaptation happens.
- Never two hard days back to back.
- If you are still tired 48 hours later, the session was too hard or you need more easy weeks.
- Niggle in a tendon or joint? Stop the intervals. Not “run through it.” That path is on my own page, and it is not a good one.
Measuring it
- Lived A proper lab test (CPET) with a mask is the real thing, and the only way to get a trustworthy number. Worth doing once for a baseline.
- Watch estimates are convenient and roughly indicative. ⚠ They are estimates from heart rate and pace, not measurements, so treat the trend as informative and the absolute number as decorative.
- The practical test: in twelve weeks, the same run at the same heart rate should be faster, or the same pace should feel easier. That is the adaptation, whatever the number says.
What people get wrong
- Starting intervals in week one. The most common and most expensive error.
- Interval one too hard. If you cannot repeat it four times, you are not training VO2 max. You are just tired.
- Too often. Once a week is enough at first. Twice is a lot. Three times is how people stop training entirely.
- Skipping the warm-up. Ten minutes. Every time.
- Doing only intervals. They work because the rest of your week is easy.
- Trusting a heart-rate formula. Effort is the more reliable instrument.
VO2 max cheat sheet
(📊 = evidence · 📖 = framework · 🔧 = my own rule)
| 📊 The 4×4 is the most-studied protocol: 4 min at 90–95% HRmax, 3 min easy, four times. ~7% VO2 max in 8 weeks, beating moderate training at matched workload | |
| ⚠ It is best-studied, not magic. The active ingredient is time near max via long intervals (≥2 min), so 6×3 or 4×8 work too. Long intervals do beat sprints (6.5% vs 3.3%) | |
| ⚠ 10–20% of people barely respond to a standard programme, and trainability is ~47% heritable. If your number does not move, that is a known distribution, not a personal failure. Judge by feel, recovery and capability too | |
| 📊 It replicates across trained, obese and cardiac populations, and consistency is what makes it trustworthy | |
| ⚠ The endpoint is a surrogate. The one randomised trial using this protocol that measured deaths was largely null. Raise it for capability, not for promised years | |
| 🔧 Earn it over ~12 weeks: strides → 4×3 moderate → 4×4 at 85–90% → the real thing | |
| 🔧 Once a week. Never consecutive days. Deload every fourth week | |
| 🔧 If interval four collapses, interval one was too hard | |
| ⚠ Get cleared first if you have any cardiac history, uncontrolled blood pressure, or years of sedentary time behind you |
Message me
If you get to week nine and the bridge sessions feel impossible, that is usually a sign the easy weeks were not easy enough. Message me and I will tell you what I would change.
Email: abhishek@terrak.ai LinkedIn: Abhishek Chaudhry
What’s changed
- 31 August 2026. Two balance corrections after applying the same scrutiny that rewrote the Zone 2 page. First, I had implied the 4×4 was uniquely correct; it is the best-studied protocol, but the active ingredient is time near maximum via long intervals, and 6×3 or 4×8 achieve it too. Second, and more importantly, I had reported “~7% in eight weeks” as though it were a personal prediction. It is a group mean. 10–20% of people show limited trainability and the response is roughly 47% heritable, which is now a section of its own, because not knowing it is how people wrongly conclude they failed.
- 31 August 2026. First published.