How accurate is Apple Watch HRV compared to a chest strap and Oura Ring?
Apple Watch underestimates HRV by ~8 ms vs a Polar H10 chest strap; Oura agrees closely with ECG overnight — but the metrics differ, so direct comparison is misleading.
On this page
TL;DR
Apple Watch, the Oura Ring, and a Polar H10 chest strap all report a number labelled "HRV," but the three are not the same measurement. Apple Watch reports SDNN from opportunistic short windows, sampled by an optical PPG sensor at the wrist. The Oura Ring reports the mean of 5-minute rMSSD samples across the whole night, also from PPG but at the finger. The Polar H10 derives HRV from an ECG signal at the chest and is the device most validation studies use as their reference. In a 2024 peer-reviewed study (O'Grady et al., Sensors), Apple Watch Series 9 and Ultra 2 underestimated HRV by an average of 8.31 ms (p = 0.025) versus a Polar H10 reference, with a mean absolute error of 20.46 ms and a mean absolute percentage error of 28.88%. Studies of the Oura Ring against ECG find that nocturnal rMSSD agrees closely with the chest reference. None of this means one device is "wrong" — they measure different things on different schedules and the metrics are not interchangeable.
Why a single number called "HRV" across three devices is misleading
Heart rate variability is the variation in the time interval between consecutive heartbeats. From that single underlying signal, researchers compute several distinct metrics — SDNN, rMSSD, low-frequency power, high-frequency power, the LF:HF ratio, and others — each of which captures a slightly different aspect of how the autonomic nervous system is modulating the heart.
Apple, Oura and Polar each choose differently:
- Apple Watch exposes HRV as SDNN — the standard deviation of normal-to-normal beat intervals — through HealthKit (<code>HKQuantityTypeIdentifierHeartRateVariabilitySDNN</code>). Apple does not expose rMSSD or frequency-domain values directly.
- Oura computes rMSSD from 5-minute windows across the night and reports the nightly average. Per Oura's own writing on the topic (accessed 15 May 2026), the choice of metric and the choice to use the entire night come from the lab of Marco Altini, who advises Oura.
- Polar H10 records the underlying R-R interval stream from an ECG-based chest strap (accessed 15 May 2026). It does not pick a single HRV metric for you; the strap delivers the R-R data and you (or your training app) compute whichever metric you want.
This matters before you compare numbers. SDNN and rMSSD answer different questions; opportunistic short windows and overnight averages capture different physiology. Even with a perfectly accurate sensor, the same person would show different absolute values across the three devices.
Apple Watch HRV versus a Polar H10 chest strap
The most directly relevant recent study is O'Grady et al., "The Validity of Apple Watch Series 9 and Ultra 2 for Serial Measurements of Heart Rate Variability and Resting Heart Rate," published in Sensors in 2024 (PMC11478500).
The study design: 39 healthy adults wore both an Apple Watch (Series 9 or Ultra 2) and a Polar H10 chest strap during a controlled 5-minute measurement each morning over 14 days, lying supine and still. The chest strap, paired with Kubios HRV analysis software, served as the reference standard. Over the cohort, 316 paired HRV measurements were collected.
The headline numbers:
- Apple Watch underestimated HRV by an average of 8.31 ms relative to the Polar H10 reference.
- The mean absolute error (MAE) was 20.46 ms.
- The mean absolute percentage error (MAPE) was 28.88%.
- For resting heart rate measured at the same time, the agreement was very close (mean difference: −0.1 bpm), and the study notes that errors in resting heart rate did not appear to drive the HRV error — meaning Apple Watch can read the heart rate well while still disagreeing with the chest strap on the HRV derived from it.
Plain-English summary: for an absolute SDNN value at a given moment, Apple Watch and a Polar H10 are not interchangeable. A ~29% mean absolute percentage error against the chest-strap reference is meaningful. The study authors note that this kind of disagreement is consistent with what is known about wrist-PPG HRV: motion artifacts, peripheral pulse-wave timing and the algorithmic step of inferring beat intervals from a continuous optical signal all add error that an ECG electrode at the chest does not see.
What Apple Watch does do well, per the same study: track resting heart rate, and show directional changes over time in HRV that are useful at the individual level — your trend over weeks is more informative than any single morning value.
Oura Ring HRV versus an ECG reference
The most-cited validation of Oura is Cao et al., "Accuracy Assessment of Oura Ring Nocturnal Heart Rate and Heart Rate Variability in Comparison With Electrocardiography in Time and Frequency Domains," published in the Journal of Medical Internet Research in 2022 (e27487).
The study design: 35 healthy participants wore an Oura Ring and a Shimmer3 medical-grade chest ECG monitor simultaneously through overnight sleep. The ring derives inter-beat intervals (IBI) from infrared PPG at the finger at a 250 Hz sample rate.
The headline findings:
- For nocturnal heart rate and rMSSD, the Oura Ring showed close agreement with the ECG reference, both in 5-minute windows and in averages computed across the entire night.
- For AVNN, pNN50, HF and SDNN, agreement was acceptable in the average-per-night test, though less reliable in 5-minute windows.
- For frequency-domain HRV — specifically LF power and the LF:HF ratio — error rates against the ECG reference were high in both tests.
A follow-up 2024 study (Liang, Yilmaz and Soon, Sensors, mdpi.com/1424-8220/24/23/7475) examined how data-quality filtering affects Oura's accuracy across 92 younger and 22 older participants. With a strict validity-proportion filter (≥80% of 5-minute windows clean) and aggregation over windows of at least 30 minutes, heart rate, rMSSD and the normalised high-frequency component (HFnu) showed very high correlations with ECG, including for older participants.
A 2025 multi-device study (Dial et al., Physiological Reports, PMC12367097) compared five wearables — Garmin Fenix 6, Oura Generation 3, Oura Generation 4, Polar Grit X Pro and WHOOP 4.0 — against an ECG reference across 536 nights of sleep. Oura Gen 3 and Gen 4 showed the strongest agreement of the five devices tested, with Lin's concordance correlation coefficients of 0.97 and 0.98 for resting heart rate, respectively. Apple Watch was not included in that study.
Plain-English summary: for overnight rMSSD and nocturnal heart rate, the Oura Ring is the closest to an ECG reference of the consumer wearables that have been head-to-head tested. For frequency-domain HRV components, Oura is materially less accurate than ECG. Like Apple Watch, Oura is best used to track your own trend over time.
Where the three devices agree and where they diverge
| Question | Apple Watch | Oura Ring | Polar H10 |
|---|---|---|---|
| Sensor type | Wrist PPG (optical) | Finger PPG (optical, infrared, 250 Hz) | Chest ECG (electrical) |
| HRV metric reported | SDNN | Average of 5-minute rMSSD across the night | Raw R-R intervals (you choose the metric) |
| Sampling window | Opportunistic short windows during the day + Mindfulness session | Entire night, in 5-minute blocks | Continuous while worn |
| Accuracy vs an ECG / chest-strap reference (recent peer-reviewed) | Underestimates SDNN by ~8 ms; MAE 20.46 ms; MAPE 28.88% (Series 9 / Ultra 2 vs Polar H10, 2024) | Nocturnal rMSSD and HR agree closely with ECG; frequency-domain HRV less accurate (2022, 2024) | Reference standard in most validation studies |
| Best use | Directional trend in everyday wellness context | Directional trend in nightly recovery context | Research, training load, situations requiring an absolute value |
A few clarifications on what this table is not saying. It is not saying Apple Watch is "less accurate than Oura" — those two studies used different references, different cohorts, different metrics and different time windows. No published study has compared all three head-to-head in the same protocol on Apple Watch Series 9/Ultra 2 / Series 11 hardware. What it is saying: each device is most useful as a longitudinal tracker against itself, and an absolute HRV number from any consumer wearable is not interchangeable with a chest-strap value at a moment in time.
What this means for you in practice
If you are deciding which device to trust as a wellness signal, the practical guidance is this:
- Pick one device and stick to it. Your trend is what matters. Switching devices mid-trend resets your baseline and obscures meaningful change.
- Treat absolute HRV values across devices as not comparable. Two devices showing 45 ms and 70 ms on the same night is normal — the metric and the window are different.
- Look at your trend over weeks, not nights. Single-night noise — sampling timing, sleep quality, alcohol the night before — can swamp the underlying signal.
- For training-load decisions where the absolute value matters, use a chest strap. This is what serious endurance athletes do, and it is what most published validation studies use as reference.
For the deeper "what causes HRV to change overnight" question, see our companion piece on Apple Watch resting heart rate and what overnight changes mean and the pillar on what Apple Watch actually measures for the full sensor stack.
Where Sam Health fits in
Your Apple Watch HRV trend is one of the most useful wellness signals it produces — but the Health app shows it as a single number with no context. Sam reads your HRV from HealthKit and helps you see your week-over-week pattern, alongside the other inputs that move it (sleep duration, training load, alcohol-flagged days).
Sam Health is a wellness app, not a medical device. Consult a healthcare professional for medical advice.
Try Sam HealthSources
- O'Grady, B., Lambe, R., Baldwin, M., Acheson, T. & Doherty, C. The Validity of Apple Watch Series 9 and Ultra 2 for Serial Measurements of Heart Rate Variability and Resting Heart Rate. Sensors 24(19): 6220, 2024. https://pmc.ncbi.nlm.nih.gov/articles/PMC11478500/ — accessed 15 May 2026.
- Cao, R., Azimi, I., Sarhaddi, F., Niela-Vilen, H., Axelin, A., Liljeberg, P. & Rahmani, A.M. Accuracy Assessment of Oura Ring Nocturnal Heart Rate and Heart Rate Variability in Comparison With Electrocardiography in Time and Frequency Domains: Comprehensive Analysis. Journal of Medical Internet Research 24(1): e27487, 2022. https://www.jmir.org/2022/1/e27487 — accessed 15 May 2026.
- Liang, T., Yilmaz, G. & Soon, C.-S. Deriving Accurate Nocturnal Heart Rate, rMSSD and Frequency HRV from the Oura Ring. Sensors 24(23): 7475, 2024. https://www.mdpi.com/1424-8220/24/23/7475 — accessed 15 May 2026.
- Dial, M. et al. Validation of nocturnal resting heart rate and heart rate variability in consumer wearables. Physiological Reports 13(15), 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12367097/ — accessed 15 May 2026.
- Apple Developer. heartRateVariabilitySDNN — HKQuantityTypeIdentifier. https://developer.apple.com/documentation/healthkit/hkquantitytypeidentifier/heartratevariabilitysdnn — accessed 15 May 2026.
- Oura. Why Oura Uses The Entire Night Of HRV Data. The Pulse Blog, March 2021 (updated July 2024). https://ouraring.com/blog/hrv-data/ — accessed 15 May 2026.
- Polar. Polar H10 Heart Rate Sensor. https://www.polar.com/us-en/sensors/h10-heart-rate-sensor — accessed 15 May 2026.
Frequently Asked Questions
Does Apple Watch measure HRV continuously?+
No. Apple Watch samples HRV opportunistically throughout the day — typically when you are still — plus a measurement during a Mindfulness breathing session. It does not continuously stream beat-to-beat intervals like a chest strap does. The metric Apple exposes through HealthKit is SDNN, the standard deviation of normal-to-normal beat intervals.
Can I compare my Apple Watch HRV value to my friend's Oura value directly?+
No. Apple Watch reports SDNN sampled in short opportunistic windows. Oura reports the average of 5-minute rMSSD samples across the entire night. SDNN and rMSSD are different mathematical operations on the same beat-interval data and produce different absolute values. Compare each device against itself over time, not across devices.
What is rMSSD and how is it different from SDNN?+
rMSSD (root mean square of successive differences) emphasises beat-to-beat differences and is widely used to estimate parasympathetic (vagal) tone. SDNN (standard deviation of normal-to-normal intervals) captures total variability over a window and is sensitive to both sympathetic and parasympathetic input. Apple Watch exposes SDNN; Oura uses rMSSD.
Is a chest strap better than a watch for HRV?+
For absolute accuracy, yes. The Polar H10 is ECG-based and is the device most validation studies use as the reference standard — Polar lists it as used in over 600 scientific studies in 2022–2023. For convenience and longitudinal trends in a wellness context, a watch or ring is sufficient. The right tool depends on what question you are answering.
Why does my Apple Watch HRV vary so much night to night?+
Three reasons. First, HRV genuinely varies day to day with sleep, training load, alcohol, illness and stress — that is the whole point of tracking it. Second, Apple's opportunistic sampling means the time-of-day of measurement is not identical from night to night. Third, motion artifacts at the wrist add noise that an ECG chest strap does not see. Focus on your trend over weeks, not on any single value.
What does the research say about WHOOP HRV accuracy?+
A 2025 multi-device study (Dial et al., Physiological Reports) compared five wearables against an ECG reference across 536 nights of sleep. WHOOP 4.0 showed moderate agreement with ECG for nocturnal HRV. Oura Generation 3 and 4 showed the strongest agreement of the five devices tested. Apple Watch was not in that study.
Should I use my HRV to make health decisions?+
No. HRV is a wellness signal and a useful input for adjusting training and recovery. It is not a clinical metric, and the consumer devices that report it are not medical devices for diagnosing autonomic dysfunction or any other condition. For medical concerns, talk to a clinician.