Independent lab tests find Apple Watch underestimates VO₂max by about 6 mL/kg/min
Is Apple Watch VO₂max accurate?
The finding
In 2025 the watch read 6.07 mL/kg/min low on average (95% CI 3.77–8.38), and in 2026 6.25 low. In the 2026 study it placed only 5 people (14%) in their correct FRIEND percentile band.
- Authors
- Lambe R, O’Grady B, Baldwin M, Doherty C; Lambe R, Schumann M, Donnelly L, et al.; Caserman P, et al.
- Published
- PLoS ONE 20(5):e0323741; Mayo Clin Proc Digit Health 4(2):100357; JMIR Biomed Eng 9:e59459 · 2025; 2026; 2024
- Evidence
- Read in the primary source; numbers checked.
- Source
- doi.org/10.1371/journal.pone.0323741
How it was studied
Three small lab studies against indirect calorimetry: 30 adults on a treadmill (2025), 40 adults with a Series 10 (2026), and 19 adults on a cycle with a Series 7 (2024).
The numbers
2025: limits of agreement −6.11 to 18.26, mean absolute error 6.92, MAPE 13.31%. 2026: limits of agreement −18.23 to 5.73, mean absolute error 6.79, MAPE 13.2%; another 12 people (34%) were one band off.
2024 (cycle): lab 45.88 ± 9.42 against watch 41.37 ± 6.5, ICC 0.47, MAPE 15.79%. The watch overestimated low-fit people and underestimated high-fit people.
What it does not show
Small samples, and the test mode differs (cycle reads lower than treadmill). A 2026 living review found only one Apple Watch VO₂max study, so there is no pooled estimate yet.
What Longstanding does with it
If your VO₂max comes from a watch, the app’s percentile can be a band or two off. We would rather say that than show it as a precise number.