How accurate is wrist heart rate during exercise, versus a chest strap?
Wrist-worn optical heart rate is reasonably accurate at rest and during steady, moderate exercise — typically within a few beats per minute of a chest strap — but the gap widens substantially as effort becomes more intense, more variable, or involves gripping or lifting. A widely cited validation study found chest straps agree with ECG at rc=.996 (essentially perfect) across treadmill, elliptical and bike testing, while wrist devices only reached rc=.67-.92 over the same conditions. The practical takeaway isn't "wrist trackers are unreliable" — it's that *how* you're exercising determines whether the number on your wrist is trustworthy, and the activities where it degrades most (intervals, rowing, lifting) are exactly the ones where people most want a precise number.
This is a different question from why HRV and resting heart rate differ between wrist, bicep and chest-strap placement, which is mostly about overnight and resting readings and about which HRV formula (RMSSD vs. SDNN) each device reports. This one is about heart rate specifically, during active exercise, where motion, cadence and grip introduce failure modes that simply don't exist while you're lying still asleep.
Why the gap between wrist and chest strap widens with effort, not just movement
A chest strap uses ECG (electrocardiography): electrodes against the skin read the heart's own electrical signal directly, the same principle used in a hospital rhythm strip. It's precise enough that essentially every validation study in this field uses a chest strap or a clinical ECG as the reference everything else is judged against.
A wrist or arm band uses PPG (photoplethysmography): it shines light into the skin and measures how much bounces back, which changes slightly with each pulse of blood. That's an indirect physical proxy for a heartbeat, and it has two disadvantages a chest strap doesn't:
- It measures a blood-volume change, not an electrical trigger, which means the signal trails the heart's actual timing by roughly a second or two. During steady effort that lag is invisible. During intervals, where your heart rate is rising or falling by double digits every 10-30 seconds, the sensor is chronically a beat behind, which shows up as understated peaks and overstated troughs.
- It can lock onto the wrong rhythm. A documented artifact sometimes called "cadence lock" happens when the algorithm's motion-compensation logic mistakes the repetitive vibration of your stride or pedal stroke for the pulse signal itself, and the displayed number snaps toward your steps- or pedal-per-minute rate. It's most likely to occur when cadence and heart rate happen to sit in a similar numeric range, and it self-corrects once the two diverge — but a reading that tracks your cadence a little too suspiciously for more than a few seconds is worth a manual pulse check.
Neither of these problems appears at rest, which is why "my watch was fine all morning" and "my watch went haywire during intervals" are both completely normal experiences with the same device.
Error by activity: what the validation studies actually found
Different exercise modes stress a wrist sensor in different ways — steady cardio mostly tests motion tolerance, intervals test lag, and rowing and lifting test grip and blood-flow changes at the wrist itself. Pulling several recent validation studies together by activity type:
| Activity | Typical wrist-vs-chest-strap error | What's driving it |
|---|---|---|
| Rest / walking | Roughly 1-4% (a few bpm) | Minimal motion, stable blood flow — the easiest condition for PPG |
| Steady-state cycling | Roughly 2-3 bpm at moderate effort; ~4-5 bpm during short sprint efforts | Mostly stationary wrist, but sustained pedal cadence raises cadence-lock risk |
| Steady-state / tempo running | Roughly 3-4 bpm at threshold pace | Arm swing adds motion artifact that scales with pace |
| High-intensity intervals / sprints | 5-8 bpm lag (mid-range devices), 8-12 bpm lag (worse-performing devices); one burpee-based study found agreement collapsing below the threshold considered acceptable | Optical lag can't keep up with rapid, repeated surges and drops |
| Rowing | Among the worst of common cardio modes in device comparisons | Repetitive gripping and wrist flexion disturb the light path more than most other steady rhythmic activity |
| Resistance / circuit training | Mean underestimation of roughly 7 bpm, with some circuit-training comparisons showing errors ranging from roughly -40 to +27 bpm in the worst cases | Grip tension changes blood flow to the hand, and set/rest pauses confuse a sensor built for continuous rhythmic motion |
The pattern across all of it: accuracy degrades progressively with intensity and unpredictability, not with movement alone. A slow, steady bike ride barely troubles a wrist sensor. A burpee circuit or a heavy set of deadlifts does, because both intensity swings and grip changes are exactly what PPG handles worst.
The skin-tone wrinkle that's easy to miss
Most of the accuracy conversation focuses on motion, but there's a second, less-discussed confound: skin tone appears to interact specifically with exercise intensity, not rest. A 2025 study comparing a wrist-worn device to a chest-strap reference found no meaningful difference in error by skin tone group while at rest, but at higher exercise intensity, error rose by roughly 7.6-11.8 bpm for medium and darker Fitzpatrick skin-tone groups relative to lighter tones on the same device. This isn't a reason to distrust every reading — it's a reason not to assume any one person's "my watch runs high during hard efforts" is purely about how hard they're pushing.
A practical protocol: when to trust the wrist, when to strap on a chest monitor
- Match the tool to the activity, not the device's marketing. For Zone 2 or other steady, conversational-pace cardio — see a 4-week Zone 2 plan for how to find that zone — wrist accuracy is genuinely good, the one context where the error tables above look almost as good as a chest strap. For anything built around hitting a specific number in real time (interval repeats, threshold intervals, heavy lifting), treat the wrist number as a rough guide and use a chest strap or arm band if the number itself matters.
- Snug the band up before hard efforts. A looser fit increases motion artifact at exactly the intensities where the sensor is already struggling most; most manufacturers recommend a firmer fit for exercise than for all-day wear.
- Watch for cadence lock specifically during rhythmic efforts. If your displayed heart rate looks suspiciously close to your running cadence or pedal rpm for more than a few seconds — especially at the start of a hard effort — check it against a brief manual pulse count or a chest strap before trusting it for pacing decisions.
- Don't chase single spikes during burpees, circuits or rowing. These are exactly the conditions where every device in the comparison studies above performs worst, wrist-worn or not — a strange number mid-set is far more likely to be sensor confusion than a real cardiac event.
- If precision genuinely matters — training zones tied to a cardiac condition, a doctor-set heart-rate limit, or serious interval work — use a chest strap as the primary reading, not the wrist device, for exactly those sessions.
When it's not a device problem
Wrist-vs-chest-strap disagreement during exercise is overwhelmingly a measurement story, not a medical one. It's worth a different conversation with a doctor if:
- A chest-strap reading itself — not just the wrist number — shows an irregular rhythm during exercise
- Your heart rate fails to rise with effort the way it used to (or rises far faster than effort would suggest), consistently, across sessions
- Exercise brings on chest discomfort, unusual breathlessness, dizziness or fainting, regardless of what any device displays
A device disagreeing with another device is a sensor story. A device agreeing with itself but showing something that doesn't match how effort should behave is a different conversation, and one worth having with a doctor rather than another wearable — see when a resting heart rate deviation itself needs medical attention for that line more generally.
Where this fits into a bigger picture
Exercise heart rate accuracy is one input into a much larger question — what your recovery actually looks like the next day, and how that compares across your own history rather than against someone else's number. Vita's recovery score is built to weight sources appropriately and flag when a reading looks like an artifact rather than a real physiological signal, so a single noisy interval session doesn't quietly distort a longer trend it was never designed to represent.
FAQ
Is a chest strap more accurate than a wrist heart rate monitor during exercise?
Yes, consistently. Chest straps read the heart's electrical signal (ECG), the same principle hospitals use, while wrist and arm bands infer heart rate optically (PPG) from blood-volume changes under the skin. A widely cited validation study found chest-strap agreement with ECG at rc=.996 (near-perfect) versus rc=.67-.92 for wrist devices across treadmill, elliptical and bike testing. The gap is small at rest and steady effort, and grows during intervals, rowing and lifting.
Why does my wrist heart rate lag behind or spike oddly during interval training?
Optical sensors measure a physical change in blood volume, which trails the heart's actual electrical trigger by roughly a second or two — enough delay that a monitor is still catching up to your last effort surge when the next one starts. Research on smartwatches during sprint intervals found the lag growing to 8-12 bpm behind a chest-strap reference, versus roughly 3-4 bpm during steady tempo effort.
What is "cadence lock" and can it really make my heart rate reading wrong?
Cadence lock is a documented optical-sensor artifact where the algorithm briefly tracks the rhythmic vibration of your stride or pedal stroke instead of the true optical pulse signal, so the displayed heart rate snaps toward your steps- or pedal-per-minute rate rather than your actual heart rate. It's most likely when the two rhythms are numerically close, and it resolves itself once the rhythms diverge again — but a reading that looks suspiciously identical to your cadence for more than a few seconds is a reason to double-check by feel or a chest strap rather than assume it's correct.
Is my wrist heart rate reliable enough for Zone 2 training specifically?
Generally yes. Zone 2 is by definition low, steady-state effort, which is exactly the condition where wrist optical sensors perform best — validation studies put steady moderate-intensity error in the low single-digit bpm range. The accuracy problem is concentrated in intervals, sprints, rowing and resistance training, not in the kind of easy, conversational-pace effort Zone 2 calls for.
Does skin tone actually affect wrist heart rate accuracy during exercise?
There's now direct evidence for it, and specifically during exercise rather than at rest. A 2025 study found no difference in error by skin tone while at rest, but at higher exercise intensity, error rose by roughly 7.6-11.8 bpm for medium and darker skin tones compared to lighter tones on the same device. This isn't universal across every wearable, but it's a real, measured confound worth knowing about rather than assuming any inaccuracy is purely about motion.
This article is general health and training reference, not medical advice — see our sources & methodology. Consult a doctor for health concerns.