Wearable Heart Rate Accuracy Leaderboard
Reviewed by Sameer Sontakey, Co-founder and Chief AI & Algorithms Officer at Sensor Bio. ChatPPG is independent and device-neutral; it is not a Sensor Bio product.
In a 50-person ECG-referenced exercise study, the Polar H7 chest strap agreed best with ECG (concordance 0.996), followed by Apple Watch (0.92). Wrist-worn optical devices were least accurate on the elliptical with arm levers.
Which wearable is most accurate for heart rate?
75 published results from 30 studies in the wearable accuracy database. Each ranking below compares devices only on the same statistic against the same kind of reference. Results from different statistics or references are never ranked together.
Evidence by brand
- Apple Watch12 studies, n = 551
- Oura5 studies, n = 181
- WHOOP5 studies, n = 123
- Garmin5 studies, n = 419
- Fitbit6 studies, n = 258
- Pixel Watch2 studies, n = 79
- Samsung4 studies, n = 175
- Polar10 studies, n = 368
- Withings2 studies, n = 29
- Amazfit1 study, n = 51
- UltrahumanNo published validation found
- COROSNo published validation found
Heart rate at rest: mean absolute percentage error (%)
Reference: ECG. Lower is better. 3 studies; a device tested in more than one study shows its median.
- 1Apple Watch Series 61.15%1 study, n = 78 (1.15%) · Bonneval et al. (2025)
- 2HRV4Training (app)2.34%1 study, n = 5 (2.34%) · Stone et al. (2021)
- 3Apple Watch Series 9 / Ultra 25.91%1 study, n = 39 (5.91% (MAE 3.73 bpm)) · O'Grady et al. (2024)
- 4CameraHRV (app)17.3%1 study, n = 5 (17.26%) · Stone et al. (2021)
Heart rate at rest: mean absolute error (bpm)
Reference: ECG. Lower is better. 1 study.
- 1Fitbit Charge 47.8 bpm1 study, n = 23 (7.8 bpm (seated rest)) · Nissen et al. (2022)
- 2Samsung Galaxy Watch Active29.4 bpm1 study, n = 23 (9.4 bpm (seated rest)) · Nissen et al. (2022)
Heart rate during exercise: mean absolute error (bpm)
Reference: ECG. Lower is better. 2 studies; a device tested in more than one study shows its median.
- 1Fitbit Charge 64.5 bpm1 study, n = 45 (4.5 bpm) · Gielen et al. (2026)
- 2Google Pixel Watch 24.9 bpm1 study, n = 45 (4.9 bpm) · Gielen et al. (2026)
- 3Samsung Galaxy Watch Active26.1 bpm1 study, n = 23 (6.1 bpm (squats)) · Nissen et al. (2022)
- 4Fitbit Charge 415.7 bpm1 study, n = 23 (15.7 bpm (squats)) · Nissen et al. (2022)
Heart rate during exercise: lin's concordance correlation
Reference: ECG. Higher is better. 1 study.
- 1Polar H7 (chest strap, ECG)0.9961 study, n = 50 (rc = 0.996) · Gillinov et al. (2017)
- 2Apple Watch0.921 study, n = 50 (rc = 0.92) · Gillinov et al. (2017)
- 3TomTom Spark Cardio0.831 study, n = 50 (rc = 0.83) · Gillinov et al. (2017)
- 4Garmin Forerunner 2350.811 study, n = 50 (rc = 0.81) · Gillinov et al. (2017)
- 5Scosche Rhythm+0.751 study, n = 50 (rc = 0.75) · Gillinov et al. (2017)
- 6Fitbit Blaze0.671 study, n = 50 (rc = 0.67) · Gillinov et al. (2017)
Overnight heart rate: intraclass correlation
Reference: ECG. Higher is better. 2 studies; a device tested in more than one study shows its median.
- 1Polar Vantage V20.9981 study, n = 29 (ICC 0.998) · Nuuttila et al. (2021)Maker funded
- 2WHOOP 3.00.991 study, n = 53 (ICC 0.99) · Miller et al. (2022)Maker funded
- 3Apple Watch Series 60.961 study, n = 53 (ICC 0.96) · Miller et al. (2022)
- 4Polar Vantage V0.931 study, n = 53 (ICC 0.93) · Miller et al. (2022)
- 5Oura Ring Gen20.851 study, n = 53 (ICC 0.85) · Miller et al. (2022)
- 6Somfit0.651 study, n = 53 (ICC 0.65) · Miller et al. (2022)
- 7Garmin Forerunner 245 Music0.411 study, n = 53 (ICC 0.41) · Miller et al. (2022)
Overnight heart rate: mean absolute error (bpm)
Reference: ECG. Lower is better. 1 study.
- 1WHOOP 3.00.7 bpm1 study, n = 53 (0.7 bpm (absolute bias)) · Miller et al. (2022)Maker funded
- 2Apple Watch Series 61.5 bpm1 study, n = 53 (1.5 bpm (absolute bias)) · Miller et al. (2022)
- 2Polar Vantage V1.5 bpm1 study, n = 53 (1.5 bpm (absolute bias)) · Miller et al. (2022)
- 4Oura Ring Gen21.8 bpm1 study, n = 53 (1.8 bpm (absolute bias)) · Miller et al. (2022)
- 5Somfit2.6 bpm1 study, n = 53 (2.6 bpm (absolute bias)) · Miller et al. (2022)
- 6Garmin Forerunner 245 Music5.4 bpm1 study, n = 53 (5.4 bpm (absolute bias)) · Miller et al. (2022)
Overnight heart rate: mean absolute percentage error (%)
Reference: ECG. Lower is better. 1 study.
- 1Oura Ring Gen31.67%1 study, n = 13 (1.67%) · Dial et al. (2025)
- 2Oura Ring 41.94%1 study, n = 13 (1.94%) · Dial et al. (2025)
- 3Polar Grit X Pro2.71%1 study, n = 13 (2.71%) · Dial et al. (2025)
- 4WHOOP 4.03%1 study, n = 13 (3.00%) · Dial et al. (2025)
Overnight heart rate: lin's concordance correlation
Reference: ECG. Higher is better. 1 study.
- 1Oura Ring 40.981 study, n = 13 (0.98) · Dial et al. (2025)
- 2Oura Ring Gen30.971 study, n = 13 (0.97) · Dial et al. (2025)
- 3WHOOP 4.00.911 study, n = 13 (0.91) · Dial et al. (2025)
- 4Polar Grit X Pro0.861 study, n = 13 (0.86) · Dial et al. (2025)
Head-to-head results within single studies
Devices tested side by side in one study, on the same people and protocol.
Ranked: higher is better
Lin's concordance correlation (rc) vs ECG, all exercise conditions
Gillinov S, et al. (2017), Medicine & Science in Sports & Exercise. doi:10.1249/MSS.0000000000001284
50 healthy adults · Reference: ECG · Treadmill, stationary bike and elliptical at rest, light, moderate and vigorous intensity
- 1Polar H7 (chest strap, ECG)Brand guide0.996
- 2Apple WatchBrand guide0.92
- 3TomTom Spark Cardio0.83
- 4Garmin Forerunner 2350.81
- 5Scosche Rhythm+ (forearm)0.75
- 6Fitbit Blaze0.67
Reported result (not ranked)
Overall heart rate error (study conclusion)
Shcherbina A, et al. (2017), Journal of Personalized Medicine. doi:10.3390/jpm7020003
60 volunteers · Reference: Continuous telemetry (ECG) · Sitting, walking, running and cycling
- •Apple WatchBrand guideLowest overall error
- •Samsung Gear S2Highest overall error
- •6 of 7 devicesMedian error < 5% while cycling
All published results
Every Heart rate result in the database, including statistics that cannot be ranked. Download the full table as CSV for all rows. From the wearable accuracy database (v1.0.0).
Heart rate at rest9 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Oura Ring (Gen2) | r = 0.90Pearson correlation | Actiheart (ECG + accelerometer) | n = 21, Free-living, up to 7 days | Henriksen et al. (2022)Independent |
| Oura Ring (Gen2) | 3%Mean absolute percentage error | Actiheart (ECG + accelerometer) | n = 21, Free-living, up to 7 days | Henriksen et al. (2022)Independent |
| HRV4Training (app) | 2.34%Mean absolute percentage error | ECG | n = 5, Resting state | Stone et al. (2021)Independent |
| CameraHRV (app) | 17.26%Mean absolute percentage error | ECG | n = 5, Resting state | Stone et al. (2021)Independent |
| WHOOP | ≤ 0.39 ± 0.38%Mean bias (signed) | ECG | Rest, 15 opportunities | Bellenger et al. (2021)Maker funded |
| Apple Watch Series 9 / Ultra 2 | 5.91% (MAE 3.73 bpm)Mean absolute percentage error | Polar H10 + Kubios HRV | n = 39, Serial morning measurements over 14 days, 316 readings | O'Grady et al. (2024)Independent |
| Apple Watch Series 6R-R intervals and BPM at rest | 1.15%Mean absolute percentage error | ECG | n = 78, Lab: rest, talking, movie, before/after walking | Bonneval et al. (2025)Independent |
| Fitbit Charge 4 | 7.8 bpm (seated rest)Mean absolute error | Holter ECG | n = 23, Rest, sedentary, low- and high-intensity phases | Nissen et al. (2022)Independent |
| Samsung Galaxy Watch Active2 | 9.4 bpm (seated rest)Mean absolute error | Holter ECG | n = 23, Rest, sedentary, low- and high-intensity phases | Nissen et al. (2022)Independent |
| Garmin Vivoactive 4 | ρc = 0.9998Lin's concordance correlation | ECG | n = 263, 30-min standardised recording | Theurl et al. (2023)Independent |
| Withings ScanWatch | MAPE < 10% in all domains except shopping (10.8%); < 5% when sedentaryStudy finding (no single number) | ECG chest strap (Polar H7/H10) | n = 7, 12-h free-living | Giggins et al. (2022)Independent |
| Fitbit Charge 4 | ICC 0.80 (COPD) / 0.79 (controls)Intraclass correlation | ECG chest strap (Polar H7/H10) | n = 55, 5 days free-living | Hermans et al. (2024)Funding not reported |
- Oura Ring (Gen2)r = 0.90
Pearson correlation · vs Actiheart (ECG + accelerometer) · n = 21, Free-living, up to 7 days
Henriksen et al. (2022) Independent
Mean absolute percentage error · vs Actiheart (ECG + accelerometer) · n = 21, Free-living, up to 7 days
Henriksen et al. (2022) Independent
- HRV4Training (app)2.34%
Mean absolute percentage error · vs ECG · n = 5, Resting state
Stone et al. (2021) Independent
- CameraHRV (app)17.26%
Mean absolute percentage error · vs ECG · n = 5, Resting state
Stone et al. (2021) Independent
- WHOOP≤ 0.39 ± 0.38%
Mean bias (signed) · vs ECG · Rest, 15 opportunities
Bellenger et al. (2021) Maker funded
- Apple Watch Series 9 / Ultra 25.91% (MAE 3.73 bpm)
Mean absolute percentage error · vs Polar H10 + Kubios HRV · n = 39, Serial morning measurements over 14 days, 316 readings
O'Grady et al. (2024) Independent
- Apple Watch Series 61.15%
Mean absolute percentage error, R-R intervals and BPM at rest · vs ECG · n = 78, Lab: rest, talking, movie, before/after walking
Bonneval et al. (2025) Independent
- Fitbit Charge 47.8 bpm (seated rest)
Mean absolute error · vs Holter ECG · n = 23, Rest, sedentary, low- and high-intensity phases
Nissen et al. (2022) Independent
- Samsung Galaxy Watch Active29.4 bpm (seated rest)
Mean absolute error · vs Holter ECG · n = 23, Rest, sedentary, low- and high-intensity phases
Nissen et al. (2022) Independent
- Garmin Vivoactive 4ρc = 0.9998
Lin's concordance correlation · vs ECG · n = 263, 30-min standardised recording
Theurl et al. (2023) Independent
- Withings ScanWatchMAPE < 10% in all domains except shopping (10.8%); < 5% when sedentary
Study finding (no single number) · vs ECG chest strap (Polar H7/H10) · n = 7, 12-h free-living
Giggins et al. (2022) Independent
- Fitbit Charge 4ICC 0.80 (COPD) / 0.79 (controls)
Intraclass correlation · vs ECG chest strap (Polar H7/H10) · n = 55, 5 days free-living
Hermans et al. (2024) Funding not reported
Heart rate during exercise18 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Apple Watch Series 6 | High agreement in squat, deadlift, cable row (rho > 0.832)Study finding (no single number) | ECG chest strap (Polar H7/H10) | n = 29, Resistance exercise (5 exercises) | Støve et al. (2022)Funding not reported |
| WHOOP 3.0 | High agreement only in squats; low in rows and burpeesStudy finding (no single number) | ECG chest strap (Polar H7/H10) | n = 29, Resistance exercise (5 exercises) | Støve et al. (2022)Funding not reported |
| Fitbit Charge 6mixed rest/stress/walking | 4.5 bpmMean absolute error | ECG chest strap (Zephyr BioHarness / Faros) | n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers | Gielen et al. (2026)Funding not reported |
| Google Pixel Watch 2mixed rest/stress/walking | 4.9 bpmMean absolute error | ECG chest strap (Zephyr BioHarness / Faros) | n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers | Gielen et al. (2026)Funding not reported |
| Fitbit Inspire 3mixed rest/stress/walking | 9–14 bpm (group range)Mean absolute error | ECG chest strap (Zephyr BioHarness / Faros) | n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers | Gielen et al. (2026)Funding not reported |
| Polar Ignite 3mixed rest/stress/walking | 9–14 bpm (group range)Mean absolute error | ECG chest strap (Zephyr BioHarness / Faros) | n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers | Gielen et al. (2026)Funding not reported |
| Polar Pacermixed rest/stress/walking | 9–14 bpm (group range)Mean absolute error | ECG chest strap (Zephyr BioHarness / Faros) | n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers | Gielen et al. (2026)Funding not reported |
| Oura Ring Gen3mixed rest/stress/walking | 9–14 bpm (group range)Mean absolute error | ECG chest strap (Zephyr BioHarness / Faros) | n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers | Gielen et al. (2026)Funding not reported |
| WHOOP 4.0 (upper arm) | Strongest agreement (with Verity Sense) in graded treadmillStudy finding (no single number) | ECG chest strap (Polar H7/H10) | n = 28, Rest, warm-up, burpees, Bruce treadmill | Moghaddam et al. (2025)Funding not reported |
| Polar Verity Sense | Strongest agreement (with WHOOP upper arm) in graded treadmillStudy finding (no single number) | ECG chest strap (Polar H7/H10) | n = 28, Rest, warm-up, burpees, Bruce treadmill | Moghaddam et al. (2025)Funding not reported |
| Fitbit Charge 4squats | 15.7 bpm (squats)Mean absolute error | Holter ECG | n = 23, Rest, sedentary, low- and high-intensity phases | Nissen et al. (2022)Independent |
| Samsung Galaxy Watch Active2squats | 6.1 bpm (squats)Mean absolute error | Holter ECG | n = 23, Rest, sedentary, low- and high-intensity phases | Nissen et al. (2022)Independent |
- Apple Watch Series 6High agreement in squat, deadlift, cable row (rho > 0.832)
Study finding (no single number) · vs ECG chest strap (Polar H7/H10) · n = 29, Resistance exercise (5 exercises)
Støve et al. (2022) Funding not reported
- WHOOP 3.0High agreement only in squats; low in rows and burpees
Study finding (no single number) · vs ECG chest strap (Polar H7/H10) · n = 29, Resistance exercise (5 exercises)
Støve et al. (2022) Funding not reported
- Fitbit Charge 64.5 bpm
Mean absolute error, mixed rest/stress/walking · vs ECG chest strap (Zephyr BioHarness / Faros) · n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers
Gielen et al. (2026) Funding not reported
- Google Pixel Watch 24.9 bpm
Mean absolute error, mixed rest/stress/walking · vs ECG chest strap (Zephyr BioHarness / Faros) · n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers
Gielen et al. (2026) Funding not reported
- Fitbit Inspire 39–14 bpm (group range)
Mean absolute error, mixed rest/stress/walking · vs ECG chest strap (Zephyr BioHarness / Faros) · n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers
Gielen et al. (2026) Funding not reported
- Polar Ignite 39–14 bpm (group range)
Mean absolute error, mixed rest/stress/walking · vs ECG chest strap (Zephyr BioHarness / Faros) · n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers
Gielen et al. (2026) Funding not reported
- Polar Pacer9–14 bpm (group range)
Mean absolute error, mixed rest/stress/walking · vs ECG chest strap (Zephyr BioHarness / Faros) · n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers
Gielen et al. (2026) Funding not reported
- Oura Ring Gen39–14 bpm (group range)
Mean absolute error, mixed rest/stress/walking · vs ECG chest strap (Zephyr BioHarness / Faros) · n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/hot/cold chambers
Gielen et al. (2026) Funding not reported
- WHOOP 4.0 (upper arm)Strongest agreement (with Verity Sense) in graded treadmill
Study finding (no single number) · vs ECG chest strap (Polar H7/H10) · n = 28, Rest, warm-up, burpees, Bruce treadmill
Moghaddam et al. (2025) Funding not reported
- Polar Verity SenseStrongest agreement (with WHOOP upper arm) in graded treadmill
Study finding (no single number) · vs ECG chest strap (Polar H7/H10) · n = 28, Rest, warm-up, burpees, Bruce treadmill
Moghaddam et al. (2025) Funding not reported
- Fitbit Charge 415.7 bpm (squats)
Mean absolute error, squats · vs Holter ECG · n = 23, Rest, sedentary, low- and high-intensity phases
Nissen et al. (2022) Independent
- Samsung Galaxy Watch Active26.1 bpm (squats)
Mean absolute error, squats · vs Holter ECG · n = 23, Rest, sedentary, low- and high-intensity phases
Nissen et al. (2022) Independent
29 more results in the database.
Overnight heart rate4 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Oura Ring Gen3 | 1.67%Mean absolute percentage error | ECG | n = 13, Home, during sleep, 536 nights | Dial et al. (2025)Independent |
| Oura Ring Gen3 | 0.97Lin's concordance correlation | ECG | n = 13, Home, during sleep, 536 nights | Dial et al. (2025)Independent |
| Oura Ring 4 | 1.94%Mean absolute percentage error | ECG | n = 13, Home, during sleep, 536 nights | Dial et al. (2025)Independent |
| Oura Ring 4 | 0.98Lin's concordance correlation | ECG | n = 13, Home, during sleep, 536 nights | Dial et al. (2025)Independent |
| Polar Grit X Pro | 2.71%Mean absolute percentage error | ECG | n = 13, Home, during sleep, 536 nights | Dial et al. (2025)Independent |
| Polar Grit X Pro | 0.86Lin's concordance correlation | ECG | n = 13, Home, during sleep, 536 nights | Dial et al. (2025)Independent |
| WHOOP 4.0 | 3.00%Mean absolute percentage error | ECG | n = 13, Home, during sleep, 536 nights | Dial et al. (2025)Independent |
| WHOOP 4.0 | 0.91Lin's concordance correlation | ECG | n = 13, Home, during sleep, 536 nights | Dial et al. (2025)Independent |
| Oura Ring (Gen1/2) | −0.63 bpm (r² 0.996)Mean bias (signed) | ECG | n = 49, Nocturnal | Kinnunen et al. (2020)Maker funded |
| Apple Watch Series 6 | ICC 0.96Intraclass correlation | ECG | n = 53, Overnight, sleep laboratory | Miller et al. (2022)Other maker funded |
| Apple Watch Series 6 | 1.5 bpm (absolute bias)Mean absolute error | ECG | n = 53, Overnight, sleep laboratory | Miller et al. (2022)Other maker funded |
| Garmin Forerunner 245 Music | ICC 0.41Intraclass correlation | ECG | n = 53, Overnight, sleep laboratory | Miller et al. (2022)Other maker funded |
- Oura Ring Gen31.67%
Mean absolute percentage error · vs ECG · n = 13, Home, during sleep, 536 nights
Dial et al. (2025) Independent
- Oura Ring Gen30.97
Lin's concordance correlation · vs ECG · n = 13, Home, during sleep, 536 nights
Dial et al. (2025) Independent
- Oura Ring 41.94%
Mean absolute percentage error · vs ECG · n = 13, Home, during sleep, 536 nights
Dial et al. (2025) Independent
- Oura Ring 40.98
Lin's concordance correlation · vs ECG · n = 13, Home, during sleep, 536 nights
Dial et al. (2025) Independent
- Polar Grit X Pro2.71%
Mean absolute percentage error · vs ECG · n = 13, Home, during sleep, 536 nights
Dial et al. (2025) Independent
- Polar Grit X Pro0.86
Lin's concordance correlation · vs ECG · n = 13, Home, during sleep, 536 nights
Dial et al. (2025) Independent
- WHOOP 4.03.00%
Mean absolute percentage error · vs ECG · n = 13, Home, during sleep, 536 nights
Dial et al. (2025) Independent
- WHOOP 4.00.91
Lin's concordance correlation · vs ECG · n = 13, Home, during sleep, 536 nights
Dial et al. (2025) Independent
- Oura Ring (Gen1/2)−0.63 bpm (r² 0.996)
Mean bias (signed) · vs ECG · n = 49, Nocturnal
Kinnunen et al. (2020) Maker funded
- Apple Watch Series 6ICC 0.96
Intraclass correlation · vs ECG · n = 53, Overnight, sleep laboratory
Miller et al. (2022) Other maker funded
- Apple Watch Series 61.5 bpm (absolute bias)
Mean absolute error · vs ECG · n = 53, Overnight, sleep laboratory
Miller et al. (2022) Other maker funded
- Garmin Forerunner 245 MusicICC 0.41
Intraclass correlation · vs ECG · n = 53, Overnight, sleep laboratory
Miller et al. (2022) Other maker funded
10 more results in the database.
Read before comparing
- Devices tested are older generations; newer models have not all been re-tested under the same protocol.
- Error was higher during walking, for males, higher BMI and darker skin tone in Shcherbina et al.
- Chest straps measure electrical activity (ECG), not optical PPG, which is why they lead.
- Rankings combine studies only when they report the same statistic against the same kind of reference, and still differ in protocol and population. Treat close values as ties.
Download the wearable accuracy table (CSV)
Every leaderboard row with study, DOI, reference standard and sample size. We'll email you the link.
FAQ
Which smartwatch is most accurate for heart rate?
In Gillinov et al. (2017), Apple Watch had the best agreement with ECG among wrist devices (rc = 0.92). Shcherbina et al. (2017) also found Apple Watch had the lowest overall heart rate error of seven wrist devices.
Are chest straps more accurate than smartwatches?
Yes in the published head-to-head data. The Polar H7 chest strap reached rc = 0.996 against ECG versus 0.67 to 0.92 for optical wrist and forearm monitors (Gillinov et al., 2017).
When is wrist heart rate least accurate?
None of the devices in Gillinov et al. were accurate on the elliptical trainer with arm levers, and Shcherbina et al. reported the highest errors during walking.