Apple Watch Series 12 Heart Rate Accuracy: What Apple's Data Shows
Updated: 2 days ago
Last Updated: October 12, 2026
Apple says the Apple Watch Series 12 has "the most accurate heart rate sensing in a wearable," and it backed that with a 1,254-person study: Series 12 beat six named competitors in 139 of 159 comparisons against a chest strap. What the study does not publish is how accurate Series 12 actually is. Apple's own 2024 numbers put the previous sensor at 89% of all-day readings within 5 bpm, but Apple gives no matching figure for Series 12, never compares it to Series 11, and no independent study exists yet. Apple's last sensor upgrade did hold up under independent testing, so the claim is plausible. For now, the evidence is Apple's own. Here is what the data shows and what it leaves open.
Want to see how other watches compare in independent studies? Try the heart rate accuracy tool, then read on for the Apple-specific story.
I have been tracking wearable heart rate validation studies for most of this year, and Apple is the one brand that has published its own accuracy numbers twice. That makes it possible to line up what changed across generations using Apple's data first, then check it against independent research.
What changed in the Series 12 heart rate sensor
Apple calls the new hardware the Health Sensing System. It ships in both Series 12 and Ultra 4, and it is the first change to the optical heart sensor wording since Series 6 in 2020.
Watches | Optical heart sensor (Apple's wording) | Background heart rate |
Original, Series 1, Series 2 | Heart rate sensor | Not published |
Series 3 | Optical heart sensor | Not published |
Series 4 | Optical heart sensor | About every 5 minutes |
Series 5, every SE model | Second-generation optical heart sensor | About every 5 minutes |
Series 6 to Series 11, Ultra 1 to 3 | Third-generation optical heart sensor | About every 5 minutes |
Series 12, Ultra 4 | Always-on optical heart sensor (Health Sensing System) | Every 5 seconds |
The practical changes Apple lists are larger, more power-efficient green LEDs, heart rate measured every 5 seconds all day, and heart rate variability (HRV) measured up to 24 times more often. Those feed two new features, Readiness and Recovery HRV, which Apple lists only for Series 12 and Ultra 4. Sources: Apple Newsroom and Apple's tech specs pages for each model.
Apple's own heart rate accuracy numbers, by generation
Apple's November 2024 white paper is the only Apple document with absolute accuracy figures. It reports the share of readings within 5 bpm of a chest strap or ECG, and it splits all-day readings by sensor generation.
Watches | All-day readings within 5 bpm | Within 10 bpm |
Series 4, Series 5, SE | 72% | 85% |
Series 6 and later, all Ultra | 89% | 97% |
Series 12, Ultra 4 | Not published | Not published |
The same paper covers workouts, though it does not say which watch models were used. Lab sessions came out higher than real-world ones.
Workout | Within 5 bpm | Reading available |
Stationary bike (lab) | 99% | 99% |
Sedentary | 98% | 99% |
Treadmill running (lab) | 96% | 97% |
Cycling (real world) | 96% | 97% |
Yoga | 92% | 96% |
HIIT | 91% | 95% |
Running (real world) | 88% | 88% |
Walking (real world) | 87% | 91% |
Two things stand out. Apple publishes where it does worse, which most brands do not. And on a real-world run, Apple reported that readings were available 88% of the time. That second number matters later in this post.
Apple Watch Series 12 heart rate accuracy: what Apple's study found
Apple's September 2026 heart rate accuracy study ran in July and August 2026 at five sites. Apple ran three of them and an outside research firm ran two, which enrolled 59% of participants.
Study detail | What Apple did |
Participants | 1,460 enrolled, 1,254 with usable paired data. 20% had darker skin (Fitzpatrick V to VI), 35% female, average age 39 |
Reference | Polar H10 chest strap |
Design | Series 12 on one wrist, one competitor on the other |
Activities | Outdoor runs, treadmill intervals, outdoor cycling, HIIT, strength training, rest, desk work, cooking, walking |
Competitors | Garmin Forerunner 970, Google Pixel Watch 4, Huawei Watch 5, Samsung Galaxy Watch 8, WHOOP 5, Oura Ring 5 |
Result | Series 12 more accurate in 139 of 159 comparisons, 18 too close to call, 2 favored a competitor |
Only losses | Pixel Watch 4 at rest, by under half a beat per minute. The Pixel Watch 4 cycling comparison was also too close to call |
That is a larger and more open comparison than any other wearable brand has published. Naming every competitor, using a stated reference device, and including darker skin tones in meaningful numbers are all to Apple's credit.
What the study does not tell you
The study reports results only as Apple's error minus each competitor's error, shown as charts. That design leaves several questions open.
No Series 12 accuracy number. There is no average error in bpm and no share of readings within 5 bpm, so Series 12 cannot sit next to the 72% and 89% above.
No comparison with Series 11. The study compares finished products against competitors, so it cannot show how much the new sensor improved on the old one, a point Empirical Health also makes.
No missing-reading rate. Apple reported availability in 2024 but not here, for any device.
No results by skin tone, even though 20% of participants had darker skin.
No HRV or sleep testing, although Readiness and Recovery HRV depend mostly on overnight HRV.
Short, supervised sessions. People taking beta blockers, people with tattoos or large moles under the watch, and people with exercise restrictions were excluded.
Apple ran and analyzed it, and it is not peer reviewed.
Do independent studies back up Apple's numbers?
For the last sensor upgrade, yes. A living systematic review pooled 22 independent Apple Watch heart rate studies (Lambe et al. 2026). The average error was close to zero for every generation. What changed was the spread of individual readings.
Watches | Studies pooled | Range most readings fell within |
Series 1 to 3 | 9 | About 22 bpm wide |
Series 4, 5, SE | 5 | About 24 bpm wide |
Series 6 and later, Ultra | 8 | About 6 bpm wide |
That matches the jump Apple reported from 72% to 89%. Recent independent lab and hospital studies on Series 7 and Series 8 show similarly tight results: Kim et al. 2023 found Series 7 within about 3 bpm either way in 44 coronary artery disease patients against a 12-lead ECG, and Mulholland et al. 2025 found Series 8 within about 2 bpm during cycling.
No independent heart rate study exists yet for Series 10, Series 11, Series 12 or any Ultra since the original. Independent validation usually trails a release by two to four years. Early reviewer tests of Series 12 look close to a chest strap: one tester measured an indoor ride averaging 109.97 bpm against the strap's 109.89, and another found correlations of 0.97 to 1.0 across cycling, running and lifting, with his Series 11 slightly better on runs.
The track record makes Apple's Series 12 claim believable. Heart rate is also just the starting point. If you want to see how your own meals, caffeine and sleep move your resting heart rate and HRV, Kygo pairs your Apple Watch data with a 20 to 30 second food log. Get Kygo on iOS or Android and see what your own data shows.
The heart rate gaps Series 12 owners are reporting
A few Series 12 owners on the MacRumors forums report heart rate cutting out when they stand up or climb stairs, and gaps of up to about five minutes during hard efforts on runs. Plenty of others, including Ultra 4 owners, say theirs works fine. Tom's Guide also saw readings grayed out at the start of runs when the watch was not confident.
This is a handful of reports, not a confirmed problem, and hiding an uncertain reading is arguably the right behavior. It matters for one reason. Accuracy is usually measured only on the readings a watch actually gives you. In an independent study of five wearables, a Garmin posted the lowest error of the group while missing about half its readings (Van Oost et al. 2025). Apple reported availability in 2024 and did not in 2026, so for now nobody can check how often Series 12 gives a reading during hard exercise.
How to get more accurate heart rate on any Apple Watch
The new sensor does not change the basics. These hold for every generation.
Wear it higher. Moving a watch from just above the wrist bone to about three finger widths up the arm cut error by 11.4 percentage points in a placement study (Vermunicht et al. 2025). That study used a Fitbit, but the effect is about sensor position, not brand.
Snug, not tight. Contact pressure affects optical readings more than many people expect. A watch that slides when you move will drop readings.
Expect worse readings when your arms move. On an elliptical, holding the moving handles roughly doubled Apple Watch error, from 3.2% to 6.5% (Gillinov et al. 2017). Lifting, rowing and racquet sports cause the same problem. The full breakdown is in our post on heart rate accuracy by activity type.
Use a chest strap when it counts. For intervals or heart rate zone training, a chest strap paired to your watch is still the most reliable option.
Independent skin tone results on Apple Watch are mixed. A small Series 9 study found much larger errors for the darkest skin group, though that group had only six people. Larger studies found no significant overall effect, but more missing readings or larger errors during activity for darker skin. Apple's new study enrolled 20% darker skin tones and did not publish results by skin tone, which would settle the question for Series 12.
Common questions
Is the Apple Watch Series 12 heart rate sensor more accurate than Series 11?
Apple has not published a direct comparison. Its 2026 study tested Series 12 against six competitors only. One independent reviewer found the two close, with Series 11 slightly better on runs.
How accurate is Apple Watch Series 12 heart rate?
Apple has not published an accuracy figure for Series 12, only that it beat six competitors in 139 of 159 comparisons. The previous sensor reached 89% of all-day readings within 5 bpm in Apple's own testing.
What is the Health Sensing System on Apple Watch?
It is Apple's name for the new heart sensor hardware in Series 12 and Ultra 4. It uses larger green LEDs, measures heart rate every 5 seconds all day, and measures HRV up to 24 times more often than before.
Is Apple Watch more accurate than Garmin, Whoop or Oura for heart rate?
In Apple's own study, Series 12 beat the Garmin Forerunner 970, WHOOP 5 and Oura Ring 5 in most comparisons. That study was run by Apple and has not been independently repeated.
Does skin tone affect Apple Watch heart rate accuracy?
The evidence is mixed. One small study found larger errors for the darkest skin, while larger studies found no significant overall effect but more missing readings for darker skin. Apple has not published Series 12 results by skin tone.
Why does my Apple Watch heart rate drop out during workouts?
The watch hides readings it is not confident in, which happens most with fast arm movement or a loose fit. Wearing it higher and snugger helps, and a chest strap is the most reliable option for hard intervals.
Does Series 12 Readiness work on older Apple Watches?
Apple lists Readiness and Recovery HRV only for Series 12 and Ultra 4, and reviewers report that watchOS 27 does not add them to older models. We cover how the score works, and why it is often blank at first, in our guide to the Apple Watch Readiness score.
The bottom line
Apple's last heart sensor upgrade, Series 6, delivered a real improvement that independent studies confirmed. Series 12 may do the same, and Apple's study is bigger and more transparent than anything its competitors have published. But Apple has not released the two numbers that would let anyone measure the upgrade: Series 12's share of readings within 5 bpm, and how often it gives a reading at all. Until those numbers or an independent study arrive, treat "most accurate" as Apple's well-supported claim rather than a settled fact. For the wider picture across brands, see how accurate your heart rate monitor really is, and for Apple's other September announcement, the Apple Watch VO2 max test.
More accurate heart rate is only useful if you know what moves it. Kygo connects your food log to your Apple Watch data so you can see which habits raise or lower your resting heart rate and HRV. Download Kygo on iOS or Android to start finding your patterns.
Key sources: Apple, Using Apple Watch to measure heart rate, calorimetry, and activity (Nov 2024, Apple's own testing); Apple Watch Heart Rate Accuracy Study (Sept 2026, Apple-run, not peer reviewed); Lambe et al. 2026 (npj Digital Medicine, living systematic review); Kim et al. 2023 (Annals of Rehabilitation Medicine, Series 7); Mulholland et al. 2025 (European Journal of Applied Physiology, Series 8, devices supplied by SlateSafety); Chase et al. 2024 (Journal of Functional Morphology and Kinesiology, Series 9); Bent et al. 2020 (npj Digital Medicine, Series 4); Van Oost et al. 2025 (Sensors); Vermunicht et al. 2025 (European Heart Journal Digital Health); Gillinov et al. 2017 (Medicine and Science in Sports and Exercise); Empirical Health (Sept 2026); reviewer tests from 9to5Mac and Tom's Guide.
A note on sources. The widely repeated "Apple Watch is 90% accurate" figure is not used here. It comes from a 2017 concordance score on first-generation hardware, which is not a percentage accuracy. A 2025 study of 260 cardiac patients on Series 8 is also excluded because its reported agreement statistics are internally impossible. Owner reports of heart rate gaps are early forum posts, not measured data.
Disclaimer: Kygo is a personal data aggregation and insights platform designed for informational purposes only. The information provided by Kygo, including correlations, patterns, and trends identified in your data, does not constitute medical advice, diagnosis, or treatment. Always consult a licensed healthcare provider with any questions regarding medical conditions.
If you have a Series 12, have you seen heart rate gaps during workouts, and which ones? I would like to hear how yours compares.