Skip to content
Wearables

Wearable Accuracy: What the Studies Show

How accurate Oura, WHOOP, Garmin and Apple Watch are for heart rate, HRV, sleep stages and VO2 max, from ECG and sleep-lab studies with sample sizes.

LLLongevity Intel editorialPublished Updated 3 min read

Compiled from published studies and manufacturer specifications, checked .

We have not tested products ourselves. How we research

Quick verdict

Wearables are good at heart rate, fair at overnight HRV and sleep timing, and weak at sleep stages, calories and VO2 max. In the one 536-night ECG comparison, Oura's HRV was closest, WHOOP 4.0 moderate and a Garmin Fenix 6 poor. Most studies test older hardware in small groups.

Talk to your doctor first. This is general information, not medical advice. Supplements and devices can interact with medicines and health conditions. We quote study protocols for context and do not recommend doses.

The short answer: consumer wearables measure heart rate well, overnight HRV and sleep timing fairly, and sleep stages, calorie burn and VO2 max poorly. The best single comparison we found put five devices against an ECG for 536 nights in 13 adults: Oura’s HRV agreed most closely, WHOOP 4.0’s moderately, a Garmin Fenix 6’s poorly. Nearly every study tested hardware that has since been replaced, in small groups, so treat these as guides to what each metric can do, not a ranking of current models. Nothing here diagnoses illness.

Metric by metric

MetricWhat studies foundStudy, size, limits
Heart rate (daytime, lab)Accurate for most brands; Apple Watch and Garmin most accurate, Fitbit tended to underestimateSystematic review, 158 publications, evidence to May 2019
Resting heart rate, overnightOura Gen 4 CCC 0.98 (1.94% error), Gen 3 0.97 (1.67%); WHOOP 4.0 0.91 (3.00%); Polar 0.86; Garmin excluded by the authorsPhysiol Rep 2025, 13 adults, 536 nights, ECG reference
Overnight HRVOura Gen 4 0.99 (5.96% error); Gen 3 0.97 (7.15%); WHOOP 4.0 0.94 (8.17%); Garmin Fenix 6 0.87 (10.52%); Polar Grit X Pro 0.82 (16.32%)Same study; authors report no competing interests
Total sleep timeWHOOP closest to the lab of three devices (Fitbit Charge 4, Garmin Vivosmart 4, WHOOP) in a review of 8 studies (-1.4 min); most wearables differed significantly from the lab in one 62-adult nightReview; Sleep Adv 2025
Sleep versus wakeAbout 86 to 89% agreement for WHOOP 3.0, Oura Gen 2 and othersSensors 2022, 53 adults
Sleep stagesKappa 0.21 to 0.53 across six devices (Fitbit, Withings, Garmin Vivosmart 4, WHOOP 4.0, Apple Watch Series 8); Apple Watch highest at 0.53Sleep Adv 2025, 62 adults, one night; all devices detected over 90% of sleep but specificity was only 29 to 52%
Sleep in older adultsOura under-read total sleep by about 75 minutes and over-read deep sleep by about 71 minutes in ages 56 to 80Sleep Adv 2026, 32 people
Calories (energy expenditure)“No brand was accurate” (the review’s words)Fuller 2020 review, above
VO2 maxApple Watch Series 10 under-read by 6.25 mL/kg/min, 13.2% error, right fitness category for 14%2026 study, 40 healthy adults

What to take from it

  • Use trends, not single numbers. Even the closest HRV device had about 6% average error against an ECG; day-to-day changes inside that band mean little.
  • Ignore stage percentages. Sleep timing and duration are reasonably tracked; deep and REM figures are not reliable enough to act on.
  • Calories and VO2 max are estimates. Direction of change over months is more defensible than the value.
  • Populations differ. In an analysis of 1,536 studies, error was greater in studies with overweight samples, and greater for heart rate and calories in studies with more women (Musa 2022); Oura’s sleep readings drifted in older adults.
  • Recovery and readiness scores are not validated here. We found no independent study showing that WHOOP Recovery, Oura Readiness or Body Battery predicts health outcomes.

Limits of this page

Sample sizes are 13 to 62 adults, mostly healthy, and the hardware is one to three generations old (WHOOP 4.0 not 5.0; Oura Gen 3 and 4 not Ring 5). Some sleep-study authors report ties to device makers: the Sensors 2022 authors belong to a research group that receives support from WHOOP (WHOOP says it was not involved in that study), as the WHOOP vs Oura comparison notes. Where an abstract gave no number, none is quoted here.

Buying decisions: WHOOP vs Oura, Garmin vs WHOOP, Oura Ring review, Apple Watch health features, and the HRV guide.

Method note

Researched from PubMed abstract records and the linked papers, checked 29 Sept 2026. Abstracts only. We did not test any device. General information, not medical advice.

Common questions

Which wearable has the most accurate HRV?

In a 2025 study of 13 adults over 536 nights against an ECG, Oura Gen 4 (concordance 0.99) and Gen 3 (0.97) were closest, WHOOP 4.0 was moderate (0.94), and Garmin Fenix 6 (0.87) and Polar Grit X Pro (0.82) were poor. It is one small study on older models, so it does not rank the current Oura Ring 5, WHOOP 5.0 or fenix 9.

Can I trust my sleep stages?

Only loosely. A 62-adult sleep-lab study of six wrist devices found Cohen's kappa of 0.21 to 0.53 (fair to moderate agreement) for sleep stages, and specificity for wake of only about 29 to 52%. Use sleep duration and timing, not stage percentages.

Sources

Data and comparisons

Related reading

Wearables

Garmin vs WHOOP

Garmin's Fenix line vs WHOOP 5.0 compared on price model, battery, sensors and the published HRV and sleep validation, with who each suits.

3 min readRead
WearablesFeatured

WHOOP vs Oura: Two-Year Cost and What Studies Show

WHOOP 5.0 and Oura Ring 5 compared on two-year cost, battery and independent HRV and sleep validation, with prices re-checked 29 Sept 2026.

5 min readRead
Wearables

Smart Rings Compared: Oura Ring 4 and 5 Cost, Accuracy

Oura Ring 4 and Ring 5 prices and membership cost, how other smart rings compare on subscriptions, and what sleep and HRV validation studies found.

6 min readRead