Can Wearables Accurately Estimate VO2 Max?

In the piece about what counts as a good VO2 max, I mentioned that the number on your watch is an estimate, not a lab result, and left it there. That caveat is doing more work than it sounds like. There’s a real gap between trusting a number as a rough trend and reading it too literally, so it’s worth answering properly: how far off is that estimate, really, and does it matter which device gave it to you?

Quick digest

• Two 2025–2026 validation studies found Apple Watch underestimates VO2 max by about 6 ml/kg/min on average, but individual readings can be off by three times that in either direction.
• A meta-analysis of 14 studies found wearables that estimate VO2 max using real heart rate and pace data during exercise are more accurate on average than ones that estimate it from resting heart rate alone.
• A study of the Fitbit Charge 2 found it overestimated VO2 max by about 2.6 ml/kg/min compared to a lab test, and overestimated more in men than in women.
• The biggest source of error usually isn’t the VO2 max formula itself, it’s how accurately the device reads your heart rate during exercise, and optical wrist sensors are measurably less reliable at high intensity.
• Researchers describe wearable VO2 max as a reasonable trend indicator, not a number precise enough for a doctor to use in place of an actual test.

What “accurate” actually means here

Two different validation studies on Apple Watch, published within the last year, give a useful, concrete picture of what wearable error looks like in practice. One study had 30 people wear an Apple Watch for 5 to 10 days to generate a VO2 max estimate, then measured their real VO2 max on a treadmill with a gas-exchange mask, the actual gold standard. The watch underestimated the lab result by about 6 ml/kg/min on average (Lambe et al., 2025). A second, larger study using the newer Apple Watch Series 10 found a similar average underestimate of about 6.25 ml/kg/min, but here’s the part that matters more than the average: individual results ranged from underestimating by as much as 18 ml/kg/min to overestimating by about 6 (Lambe et al., 2026).

Put concretely: if your real VO2 max is 40, the average person’s watch might read closer to 34. But depending on the day and the person, some readings in that same study landed anywhere from 22 to 46 and still fell within the range the researchers measured. That’s the honest range of error you’re working with, not a rounding difference.

Why some wearables land closer than others

Not every device makes the same kind of estimate, and that difference matters more than which logo is on the watch. The largest meta-analysis on this question pooled 14 validation studies and split wearables into two categories: ones that estimate VO2 max using your actual heart rate and pace while you’re moving, and ones that estimate it from your resting heart rate plus basic demographics, with no exercise involved at all. The exercise-based devices, the category Garmin’s Firstbeat-powered watches fall into, showed close to zero average bias (0.09 ml/kg/min) across the pooled studies. The resting-based devices systematically overestimated fitness by about 2.17 ml/kg/min on average (Molina-García et al., 2022). Even the better category still carried real random error, with individual results scattered roughly ±9.83 ml/kg/min around the true value. A more recent systematic review of 13 studies reached a similar conclusion: exercise-based estimates track a person’s real fitness more closely, but no consumer device has been validated closely enough to replace a lab test (Železnik Mežan, 2025).

Individual devices bear this out. A study of the Fitbit Charge 2 found its “cardio fitness score” overestimated lab-measured VO2 max by about 2.6 ml/kg/min on average, with a fairly wide range from person to person (Freeberg et al., 2019). A separate study of the Apple Watch Series 7 found a mean absolute percentage error of nearly 16%, and specifically noted the watch tended to overestimate fitness in people who scored poorly on the lab test and underestimate it in people who scored well (Caserman et al., 2024). That last pattern is worth sitting with if you’re already quite fit: your watch may be more likely to undersell you, not flatter you.

What’s actually driving the error

The formula behind these estimates isn’t the weak link. All of them ultimately rely on the well-established relationship between heart rate, pace, and oxygen consumption. What breaks down is the heart rate data feeding into that formula. Optical wrist sensors, the kind built into nearly every consumer wearable, are demonstrably less accurate than a chest strap once you’re working hard, and accuracy also varies with skin tone, wrist movement, and how snugly the device fits, based on research on wrist-worn sensors across a demographically diverse group of people (Shcherbina et al., 2017). A VO2 max estimate is only as good as the heart rate reading it’s built on, so an unreliable pulse signal during a run turns into an unreliable fitness number, regardless of how good the underlying formula is.

Sex is a smaller but real factor in that same error. In the Fitbit study above, the device overestimated VO2 max more in men (by about 3.9 ml/kg/min on average) than in women (about 0.9 ml/kg/min) (Freeberg et al., 2019). That’s not a sign the device works better for women; it’s a byproduct of how the algorithm’s heart-rate-to-pace assumptions interact with the more variable muscle mass and stride mechanics in a mixed group of men. It’s one more reason to read your own number against your own trend rather than treating it as an absolute, comparable figure across anyone, of any sex, wearing a different watch.

What to actually do with this

Treat a single reading as noise and the trend as signal. One VO2 max number from your watch carries real error. Ten of them over three months, all measured the same way, tell you something a single reading can’t.

Use a chest strap if the number matters to you. Every study above measured a wrist-worn optical sensor. I wear a chest strap for the specific runs my watch uses to estimate VO2 max, since it fixes the input the whole calculation depends on and removes a meaningful chunk of the error.

Don’t compare your watch’s number directly to lab-based percentile charts. The reference tables researchers use for “what’s a good VO2 max for your age” come from actual treadmill tests, not wearable estimates. Lining up a device estimate against those percentiles adds a second layer of error on top of the first.

Expect your number to shift if you switch devices. Moving from a Fitbit to a Garmin to an Apple Watch can change your reported VO2 max by several points overnight, for reasons that have nothing to do with your actual fitness. That’s an algorithm difference, not a real change in you.

If a genuinely low score is showing up, don’t stop at the watch. These devices aren’t validated for clinical risk assessment. A consistently low reading is a reasonable prompt to get an actual test, not a diagnosis to act on by itself.

FAQ

Which wearable brand estimates VO2 max most accurately? Devices that use real heart rate and pace data during exercise, which includes most Garmin watches through their Firstbeat-based algorithm, have shown smaller average error than devices that estimate fitness from resting heart rate alone in pooled research (Molina-García et al., 2022). That’s a category-level pattern, not a guarantee for any single device or person.

Why did my VO2 max jump or drop when I got a new watch? Different brands use different formulas, different heart rate sensors, and sometimes different exercise protocols to generate the number. A change of several points after switching devices usually reflects the algorithm, not a real shift in your fitness.

If the number itself isn’t precise, is it even worth tracking? Yes, as a trend. Researchers who study this consistently describe wearable VO2 max as reasonable for spotting a change in your own trajectory over months, even while cautioning against treating any single reading as exact (Železnik Mežan, 2025).

Does a chest strap actually make a measurable difference? Yes. Since the heart rate signal is the main source of error in these estimates, and chest straps read heart rate more reliably than optical wrist sensors during hard efforts, pairing one with your watch during a VO2 max-generating run reduces a real source of noise.

Is my Apple Watch or Garmin’s VO2 max reliable enough for a doctor to use? No, and the researchers behind these validation studies are explicit about that. The random error in individual estimates is too wide for clinical decision-making. If cardiorespiratory fitness needs to inform an actual medical decision, that calls for a supervised test, not a wearable reading.

My watch’s number feels too low compared to how fit I feel. Am I wrong, or is it? Possibly neither. Research on the Apple Watch Series 7 found it tends to underestimate fitness specifically in people who test well in the lab, so a number that feels low relative to your actual performance is a documented pattern, not necessarily a sign anything’s wrong with you or the device (Caserman et al., 2024).

Sources

Caserman, P., Yum, S., Göbel, S., Reif, A., & Matura, S. (2024). Assessing the accuracy of smartwatch-based estimation of maximum oxygen uptake using the Apple Watch Series 7: Validation study. JMIR Biomedical Engineering, 9, e59459. https://doi.org/10.2196/59459

Freeberg, K. A., Baughman, B. R., Vickey, T., Sullivan, J. A., & Sawyer, B. J. (2019). Assessing the ability of the Fitbit Charge 2 to accurately predict VO2max. mHealth, 5, Article 39. https://doi.org/10.21037/mhealth.2019.09.07

Lambe, R., O’Grady, B., Baldwin, M., & Doherty, C. (2025). Investigating the accuracy of Apple Watch VO2 max measurements: A validation study. PLOS ONE, 20(5), e0323741. https://doi.org/10.1371/journal.pone.0323741

Lambe, R., Schumann, M., Donnelly, L., Hamilton, S., & Doherty, C. (2026). Accuracy of VO2 max estimates from Apple Watch Series 10. Mayo Clinic Proceedings: Digital Health, 4(2), 100357. https://doi.org/10.1016/j.mcpdig.2026.100357

Molina-García, P., Notbohm, H. L., Schumann, M., Argent, R., Hetherington-Rauth, M., Stang, J., Bloch, W., Cheng, S., Ekelund, U., Ortega, F. B., & Sardinha, L. B. (2022). Validity of estimating the maximal oxygen consumption by consumer wearables: A systematic review with meta-analysis and expert statement of the INTERLIVE network. Sports Medicine, 52(7), 1577–1597. https://doi.org/10.1007/s40279-021-01639-y

Shcherbina, A., Mattsson, C. M., Waggott, D., Salisbury, H., Christle, J. W., Hastie, T., Wheeler, M. T., & Ashley, E. A. (2017). Accuracy in wrist-worn, sensor-based measurements of heart rate and energy expenditure in a diverse cohort. Journal of Personalized Medicine, 7(2), Article 3. https://doi.org/10.3390/jpm7020003

Železnik Mežan, L. (2025). Accuracy of wearables for determining the maximal oxygen uptake and lactate threshold: A qualitative systematic review. Frontiers in Sports and Active Living, 7, Article 1707991. https://doi.org/10.3389/fspor.2025.1707991

A note from us We’re Martin and Marina, Astrea’s co-founders. We’re both into data and serious about our own training, but neither of us is a doctor or a clinical researcher. The health and physiology claims in this article come from published, peer-reviewed research, not our own expertise, which is why every article ends with a Sources list above. If a claim doesn’t trace back to a real source, we cut it before it gets published.