Zone 2 and VO2 max: what the evidence says, how to measure them, and a weekly template

This page replaces a short 2012 “heart and exercise” post. It was rewritten from scratch on 09/28/2026 and fact-checked on 09/29/2026 around three questions: does VO2 max matter outside racing, is zone 2 the right way to train it, and can a watch measure it. Every study result links to the paper or brand page it came from, opened on one of those dates; the training numbers in the template are practical defaults, not study findings.

What VO2 max is and why it predicts lifespan

VO2 max is the largest amount of oxygen your body can use per minute during hard exercise, in milliliters per kilogram of body weight per minute (ml/kg/min). Clinics often express it in METs, where 1 MET is resting metabolism, using 3.5 ml/kg/min per MET.

It matters beyond sport because of the mortality data. In a Cleveland Clinic cohort of 122,007 patients followed for a median of 8.4 years, Mandsager and colleagues ranked everyone by treadmill performance against their age and sex peers. The “elite” group had an adjusted hazard ratio of 0.20 versus the “low” group, which the authors express the other way around as about a five-fold higher risk of death for the least fit. There was no plateau: elite performers still did better than merely high performers (HR 0.77), so the authors reported no observed upper limit of benefit.

A 2009 meta-analysis of 33 cohort studies (102,980 participants for the mortality analysis) put a number on each step. Every 1 MET higher maximal capacity, which the authors equate to about 1 km/h of extra running speed, was associated with 13 percent lower all-cause mortality (RR 0.87) and 15 percent fewer coronary or cardiovascular events (RR 0.85). Low fitness (under 7.9 METs) versus high fitness (10.9 METs and up) carried an RR of 1.70 for death.

Study	Population	Comparison	Result
Mandsager 2018, JAMA Netw Open	122,007 treadmill patients, median 8.4 yr follow-up	Elite vs low fitness (age/sex percentiles)	Adjusted HR 0.20 (95% CI 0.16 to 0.24)
Mandsager 2018	same	Elite vs high fitness	Adjusted HR 0.77 (95% CI 0.63 to 0.95)
Kodama 2009, JAMA	33 cohorts, 102,980 healthy adults	Per 1 MET higher capacity	RR 0.87 for all-cause mortality
Kodama 2009	same	Low (under 7.9 METs) vs high (10.9 METs or more)	RR 1.70 for all-cause mortality

Two cautions: these are observational associations (healthier people can also exercise harder), and the Cleveland Clinic patients were referred for treadmill tests, not a random slice of the public. Neither changes the direction of the finding.

What “zone 2” actually means

Zone 2 is shorthand for steady exercise below the lactate threshold, the intensity where lactate starts accumulating faster than the body clears it. That is the definition the 2025 Sports Medicine review uses. Most people approximate it with a watch heart-rate ceiling or the rule of thumb that you can still speak in full sentences. Neither proxy is precise: a lab lactate test is the only way to locate your threshold, and the zones a watch prints are formula defaults unless you set them. In the Norwegian trial below, the groups ran at 70 percent of maximal heart rate (long slow distance), 85 percent (lactate threshold) or 90 to 95 percent (intervals); popular “zone 2” sits around the first.

Is zone 2 the best way to train for the general public?

This is where internet advice and the exercise-science literature split. The popular case for zone 2 comes from elite endurance athletes, who do most of their large volumes at low intensity and have excellent mitochondrial and fat-oxidation capacity. The 2025 narrative review by Storoschuk, Moran-MacDonald, Gibala and Gurd examined whether that observation justifies a public recommendation and concluded that it does not: current evidence does not support zone 2 as the optimal intensity for improving mitochondrial or fat-oxidative capacity, and prioritizing intensities above zone 2 appears critical for cardiometabolic benefit, especially when total training volume is low.

The volume point is the practical one. A professional cyclist training 20 or more hours a week can keep most of it easy and still log hours of hard work. Someone with three to four hours a week who copies that split ends up with well under an hour of hard training, and the review argues that at low volume the hard minutes are the ones that matter most. One disclosure from the paper: an author advises and holds equity in an exercise-related company; the conclusions rest on the cited trials.

None of this makes easy training useless. It is easier to recover from, easier to do more of, and it raises VO2 max compared with not training. The disagreement is about the word “optimal.”

Intervals vs steady state: what the trials show

The cleanest comparison is Helgerud and colleagues (2007), who randomized 40 moderately trained men to four protocols matched for total oxygen consumed, three days a week for eight weeks.

Group (matched total work)	Intensity	VO2 max change
Long slow distance	70% HRmax	No significant change
Lactate threshold	85% HRmax	No significant change
15 s on / 15 s off intervals	90 to 95% HRmax, 70% recovery	+5.5% (60.5 to 64.4 ml/kg/min)
4 x 4 min intervals, 3 min recovery	90 to 95% HRmax, 70% recovery	+7.2% (55.5 to 60.4 ml/kg/min)

Stroke volume rose about 10 percent in the interval groups, which the authors link to the VO2 max gain. Context: these were already-fit men starting near 60 ml/kg/min, where easy running has little left to add.

Across a broader population, the Milanović 2015 meta-analysis of 28 controlled trials (723 healthy adults aged 18 to 45, mean baseline 40.8 ml/kg/min) found endurance training raised VO2 max by 4.9 ml/kg/min versus non-exercising controls, interval training by 5.5, and intervals beat endurance training head to head by a small margin of 1.2 ml/kg/min. Both effects were larger in people with lower starting fitness and in longer programs. Put against the Kodama numbers, a 5 ml/kg/min gain is roughly 1.4 METs, which is the difference the mortality curve cares about.

The takeaway is not “only do intervals.” Both work, intervals work slightly better in head-to-head comparisons, and the biggest gains go to the least fit, the people popular zone 2 advice tells to stay easy.

How watches estimate VO2 max, and how accurate they are

No wearable measures oxygen. Each brand fits a model to signals it can read:

• Garmin states on its VO2 max technology page that its Garmin Human Performance Lab engine estimates VO2 max during a run from the relationship between pace (external workload) and how hard your body is working (internal workload), selecting only clean segments so you do not need a formal test protocol. Recent devices also adjust for heat, humidity and altitude when paired with a phone, and select watches feed accelerometer data in for trail terrain.
• Oura calls it Cardio Capacity. Per its support article, the first number is a population estimate from your age, sex, height and weight; the app then prompts a six-minute outdoor walking test that uses GPS distance and heart rate (it cannot be done on a treadmill), and you can also enter a lab or other-device value manually. Oura suggests repeating the walk about monthly. Gen3 or later and an active membership are required.
• WHOOP lists VO2 max as one of the inputs to its Healthspan feature, alongside resting heart rate, strength activity time and lean body mass. WHOOP’s VO2 max explainer describes three tiers: a passive estimate from resting physiology, activity, sleep and your age, sex, height and weight (available after 14 recoveries in 21 days); a GPS-augmented estimate that adds pace and heart rate from an outdoor run of 15 minutes or more; and a calibrated estimate anchored to a lab value you enter. WHOOP’s own internal testing on 248 lab tests, which has not been independently published, reports a mean absolute error of 3.3 to 3.7 ml/kg/min.

How close do these get to a metabolic cart?

Study	Device	Reference	Result
INTERLIVE meta-analysis, 2022 (14 studies)	Wearables using exercise data	Lab VO2 max	Bias -0.09 ml/kg/min, limits of agreement -9.92 to +9.74
INTERLIVE 2022	Wearables using resting-condition data only	Lab VO2 max	Bias +2.17 ml/kg/min, limits -13.07 to +17.41
Carrier 2025	Garmin fenix 6, outdoor run, n=19 general population	Graded exercise test	MAPE 7.05%, concordance 0.73 (met the authors’ validity cutoffs)
Passler 2019	Garmin Forerunner 920XT, n=24	Indirect calorimetry	MAPE 7.3%, significant underestimation; other trackers over 10%
Caserman 2024	Apple Watch Series 7, n=19	Cycle ergometer test	MAPE 15.79%, ICC 0.47; underestimated fit users, overestimated unfit ones

Reading the table: at the population level, exercise-based estimates are essentially unbiased. For one person, the plus or minus 10 ml/kg/min agreement band is almost 3 METs, wider than Kodama’s entire intermediate fitness band (7.9 to 10.8 METs), and the INTERLIVE authors say plainly that for sport or clinical purposes these methods still need improvement.

Practical rules that follow:

1. Compare only readings from the same device, activity type and conditions. A run in Puerto Rico’s afternoon humidity and one at dawn are different tests, which is why Garmin models heat.
2. Ignore single-day changes. Look at the 4 to 8 week direction.
3. If the number matters (a health decision, a race goal), get one lab test and track the watch trend from that anchor. Oura and WHOOP both let you type the lab value in.

A 3-session weekly template

This template meets the ACSM position stand minimums (150 moderate or 75 vigorous cardio minutes per week, or a mix, plus resistance work on 2 to 3 days) and includes the interval stimulus the trials above credit for VO2 max. Times are the cardio portion only; add strength days around them.

Day	Session	Intensity	Time	What it is for
1	4 x 4 minutes hard, 3 minutes easy between	90 to 95% HRmax on the work bouts (Helgerud protocol)	10 min warm-up + 25 min + cool-down, about 45 min	VO2 max and stroke volume
2	Steady state, below lactate threshold	Zone 2 by heart rate or the talk test	45 to 60 min	Volume, recovery between hard days, aerobic base
3	Moderate continuous or a second, shorter interval day	75 to 85% HRmax steady, or 15/15 intervals if recovered	30 to 40 min	Fills the ACSM moderate minutes; optional second hard stimulus

Notes on running it:

• For the first two weeks, cap the intervals at 3 x 4 minutes and let pace find the heart rate. The Helgerud subjects were moderately trained men; the same protocol is a harder relative session for a deconditioned 50-year-old.
• If you only have two days, keep Day 1 and Day 2. The review evidence says the hard day is the one not to drop when volume is low.
• Any activity works for the intervals if you can hold 90 to 95 percent of maximal heart rate: running, rowing, cycling, stairs. Watches estimate VO2 max most reliably from GPS runs, so keep one weekly run if you want a comparable number.
• Anyone with heart disease, symptoms on exertion, or years of inactivity should have a clinician sign off before interval work. The ACSM does not recommend universal exercise testing first, but does recommend it when clinically indicated.

Where to buy / how to save

The three merchants on this page are wearables that produce a VO2 max estimate. I have been a gym member since the 1990s and have run this site since 2011; the notes below are about what each device is good for, not a ranking.

• Garmin: the estimate comes from real runs with no test protocol, with heat and altitude correction and the strongest published validation record above (two independent studies under 8 percent error, though the 2019 study’s authors still judged the estimate not accurate enough for sport or health-care use). Best for runners and anyone who will do the interval sessions with GPS on.
• Oura: a ring, for people who will not wear a watch while training. The estimate comes from a monthly walking test or a manually entered lab value rather than continuous run data, and it needs an active membership.
• WHOOP: subscription-based, and folds VO2 max into Healthspan (WHOOP Age and Pace of Aging) alongside eight other metrics, including sleep, daily steps, heart-rate zone time and resting heart rate. Best for people who want VO2 max as one input rather than the headline number.

Promotions and referral offers for each are checked on their merchant pages and dated at the time of the check; a page with no offer means none was found on the merchant’s own site that day.