How accurate is your watch's VO2 max estimate?

Lab test vs. watch estimate

The gold standard is a cardiopulmonary exercise test (CPET). You run on a treadmill or ride a bike while wearing a mask that measures the oxygen you breathe in and out. The effort ramps up until you can't go any harder. That's what the norms behind our calculator come from.

A watch can't measure oxygen. Instead, it watches how hard your heart works (heart rate) at a given pace or effort (GPS speed, motion), then combines that with your age, sex, height and weight to predict what a lab would find. The general idea: if your heart rate stays lower at the same pace, you're probably fitter.

That works reasonably well, but every step adds a little room for error: the heart rate reading, the GPS pace, the max heart rate the watch assumes for you, and the terrain and weather on the day.

What the studies found, brand by brand

Most validation studies are small, and many test young, fit volunteers rather than adults over 40. Keep that in mind, and give more weight to independent studies than to company white papers.

The big picture: a 2022 meta-analysis

Researchers pooled 14 validation studies (403 people). Watches that estimate from exercise were almost spot-on as a group average (off by about 0.1 mL/kg/min). For individuals, though, the 95% range of error ran roughly ±10 mL/kg/min. Watches that estimate from resting data overestimated by about 2.2 mL/kg/min on average, with an even wider individual range (about −13 to +17). The authors concluded these estimates work at the population level but still need improvement for individual sport or clinical use.¹

Garmin (Firstbeat algorithm)

• Independent: in 19 healthy adults (average age about 25), a Garmin fēnix 6 estimate from a guided outdoor run was within about 7% of the lab value on average (mean absolute percentage error 7.05%).⁴ A similar study in 21 athletes found about 6.9%.⁵
• Over time: in 47 Air Force personnel tracked through a 12-week training program, the Garmin estimate was off by 9.4% on average. Lab VO2 max rose by about 3.2 mL/kg/min, but the Garmin number didn't change in step.⁶
• Middle-aged non-athletes: a 2025 preprint followed 203 adults (median age 55, mostly women) wearing a Garmin vívoactive 4S in everyday life for three months. The watch overestimated by about 14.5 mL/kg/min on average and was no more accurate than a simple age/sex/height/weight formula, especially for people whose fitness was lower than expected. It hasn't been peer reviewed yet.⁸
• Company data: Firstbeat, whose method Garmin uses, reports about 5% error for running and about 6% for walking in its own testing. It also shows that a wrong max heart rate adds error: a max heart rate set 15 beats too low pushed the error to about 9%.⁹

Apple Watch (Cardio Fitness)

• Independent: in 28 adults (average age about 32, mostly very fit), Apple Watch underestimated lab VO2 max by about 6 mL/kg/min on average, about 13% error.² A fitter-than-average group may partly explain the underestimate.
• Company data: Apple's validation group had 221 people with an average age of 55. It reported an average error of 1.4 ± 4.7 mL/kg/min, but against a reference projected from submaximal lab tests, not all-out tests.³
• Apple Watch estimates from Outdoor Walk, Outdoor Run and Hiking workouts. Indoor workouts don't count.¹²

Polar

• Polar has two different approaches. The Fitness Test is a 5-minute test lying down, based on resting heart rate data. Running Index is calculated from heart rate and speed on runs of at least 12 minutes. Some newer models also offer a 30-minute Walking Test.¹³¹⁵
• Independent: in 24 adults (average age 57) with cardiovascular risk factors, the resting Fitness Test averaged 13.7% error, with individual results off by up to about ±11 mL/kg/min.⁷ In the 12-week Air Force study above, it averaged 12% error and didn't pick up the real fitness gains.⁶

COROS

COROS estimates VO2 max from outdoor runs, using the relationship between your pace and heart rate, and adjusts it as you log more runs.¹⁴ We couldn't find an independent, peer-reviewed lab validation of COROS's VO2 max estimate, so we can't put a number on its accuracy yet.

WHOOP

WHOOP (a screenless strap, not a watch) now estimates VO2 max from overnight and all-day data, sharpened by GPS-tracked outdoor runs. It also lets you enter a lab result to anchor future estimates. In WHOOP's own testing against 248 lab tests, the average error was 3.3–3.7 mL/kg/min.¹⁰ That's company data. We haven't found an independent study yet.

What the error means for your percentile

Take a 45-year-old man whose watch says 35.3. On the FRIEND lab norms, that's right at the 50th percentile. Here's how much the percentile moves if the watch is off:

So a watch reading can honestly place you anywhere from "fair" to "good." That's why our calculator labels watch-based percentiles as approximate. You can check your own age group on the men 40–49 or women 40–49 norm pages, or any group in the norms library.

Why your number drifts

• Heart rate glitches. Wrist sensors can lose the signal or lock onto your running cadence, especially with a loose band, cold hands or lots of wrist movement. One bad reading can drag the estimate.
• The wrong max heart rate. Most watches start from an age formula, and your real max can be quite different. In Firstbeat's data, a max heart rate that was 15 beats off added several percentage points of error.⁹
• Heat, hills, wind and fatigue. Anything that raises your heart rate at a given pace (a hot day, dehydration, caffeine, poor sleep, a hilly route, a headwind) can nudge the number down. Polar and Apple each list several of these.¹³³
• Carrying weight or rough ground. A loaded ruck, a child carrier, a stroller, sand or snow all add work the watch can't see. Apple notes these can pull estimates low.³ If you ruck, don't expect those sessions to give you a clean VO2 max.
• Heart-rate-lowering medication. Some medications (such as beta blockers) blunt how high your heart rate goes, which can make estimates read high. Apple Watch lets you note this in the Health app.³¹² If you take a medication like this, ask your doctor how to interpret your numbers.
• Algorithms smooth on purpose. Most watches average over weeks, so a real change can take a while to show up. In one 12-week study, the watches didn't show it at all.⁶

How to get a better estimate

1. Get your profile right. Enter your correct age, sex, height and weight, and update your weight when it changes.
2. Use the right workout. Garmin wants an outdoor run with GPS, at least 10 minutes at 70%+ of your max heart rate.¹¹ Apple uses Outdoor Walk, Outdoor Run or Hiking on fairly flat ground, with a solid GPS and heart rate signal.³ Polar's Running Index needs runs of 12+ minutes at 6 km/h (3.7 mph) or faster.¹³
3. Pick a flat, familiar loop on a mild day, and give GPS a minute to lock on before you start.
4. Fix the heart rate input. Wear the watch snug, a finger's width above the wrist bone. For the cleanest data, pair a chest strap. Garmin's and Polar's watches accept one, and Garmin's manual allows either wrist or chest heart rate for VO2 max.¹¹ In the meta-analysis, chest-strap heart rate made a clear difference for resting tests, though less so for exercise-based ones.¹
5. Set your real max heart rate if you know it from a recent hard race or a supervised test. Don't go hunting for a max on your own. Talk to your doctor first if you have any heart concerns.
6. Give it a few weeks. Garmin says the first estimates "may seem inaccurate" until it has learned from a few runs.¹¹ Apple needs several outdoor workouts before its first estimate.¹²
7. If you've had a lab test, use it. WHOOP lets you enter it as an anchor.¹⁰ Either way, it tells you how far off your device runs for you.

The most useful habit: retest on the same device

Because each brand's algorithm has its own bias, switching watches can shift your number by several points even if your fitness hasn't changed. To track real progress:

• Use the same watch, the same heart rate source (wrist or strap), and similar routes and conditions.
• Log the number about every 90 days, not every day, and look at the trend over months.
• Treat a change of a point or two as noise. A steady climb over several check-ins is the signal.
• If you switch devices, start a new baseline instead of comparing old and new numbers.

Ready? Enter your watch's VO2 max in the calculator to see your approximate percentile. Then check out simple ways to raise it. Shopping for a watch? Read our best watches for tracking VO2 max after 40.

Not a diagnosis. Strong After 40 compares your numbers to published population norms for education and motivation only. It is not medical advice. Consult your doctor before starting a new exercise program or doing maximal-effort tests, especially if you have heart, joint, or other health conditions. Full disclaimer