I’ve been using WHOOP for over three years, and I can fully confirm that the heart-rate tracking from the wrist is absolutely inaccurate during running.
I compare my HR in real time with a Polar H10, a Garmin, and an Apple Watch — all of them consistently agree when I’m in Zone 2. WHOOP, instead, shows me at 165 bpm when I’m actually between 100–120 bpm. And it does this constantly.
Before anyone asks: yes, I wear the band correctly — tight, one finger above the wrist bone, fully flush to the skin, with no red or green light leaking. I’ve followed every recommendation, and after three years of daily use I know the device extremely well. So the issue is definitely not how I’m wearing it.
On a stationary bike WHOOP performs reasonably well, but during treadmill or outdoor running it becomes totally unreliable.
After 13 marathons, I know exactly what HR values to expect — and unfortunately WHOOP can’t be the device that accompanies me in training or racing.
I’ve already asked WHOOP to allow external HR devices (Polar, Garmin chest straps, or even using Apple Watch data) to feed heart-rate readings into the platform — that would completely solve the problem.
I have experienced exactly the same. Whoop 5.0 mesures my heart rate too high when I run. I believe it captures my cadence (160-ish) for 15-20 min until my I am warmed up properly. After that it is Ok but not really accurate. I must say I have the same problem with my sportswatch Polar V2. But I connect an external chest strap (armband OH1) and it is really consistent when compared to same exertion and wearing my Polar H10. I have tried it all with Whoop, wear i tight on my wrist or my biceps. It doesn’t matter, it is still inconsistent. I have less problems when doing others sports. If you find a solution for running, please send me a message.
Same for me. The first 10-15 minutes show my HR at 160+ and then it drops down quickly and closely matches the heart rate strap I wear with my Garmin watch. It does not matter if it is an outdoor run or a treadmill run. Should I be wearing the sensor somewhere else on my body for more accurate results? Currently using wrist band 24/7.
Some get better result worn on the bicep. Not me. For me the best way to wear whoop is with three fingers width from the wrist bone. It is not perfect but the best way for me.
I have exactly the same here. And it’s really a big big downside with Whoop. Specially since running is my main activity.
Whereas I also have a oura ring and the heart rate while I run is really accurate (if I compare it to my Garmin that I wear while running).
I wonder if it would be better wearing it to the biceps but it seems to be the same result for you guys.
I feel like this king of object is totally pointless if they are not even able to get the right intensity from your run since it will affect all your other figures (estimated age, etc)
I’m training and racing Surfski and K1 and have the same issue. Doing a steady Z2 session gives erratic Z3 and Z4 bursts. Going all out (Z5) at a race most of the time it was sitting in Z2 and Z3. I also use a PolarH10 as my HR device and if I just could import the HR data from it to Whoop for my training and races. Please add that in your code. Or re-design the sensor to at least be as good as Apple’s.
It’s not always too high. I have both high and low readings and zero way to correct it. I run a Garmin chest strap always and only really trust that but it screws up the strain readings on whoop.
You wonder why they behave like this as an organisation. They know full well that wrist or bicep based HR are deeply flawed approaches for active sports and somehow they refuse to improve it all while claiming they are for athletes. It’s idiotic. A whoop based chest strap would easily solve the problem but their CEO seems completely deaf to the idea. I know several people have tried to tell him directly.
I have had the same experience with an erratic and high heart rate for the first 10–15 minutes of a run,
For me, I realized my Whoop only became accurate once I started to sweat. So I started to wet my wrist and the underside of the sensor before heading out. Moistening the skin early prevented the sensor from sliding on my dry wrist and helped it get a solid reading immediately. Since I started doing this, my wrist readings have been spot-on.
I have tried wearing bicep band really tight and my readings has improved, although not perfect for skierg och skiing. But much better for slow running.
I had problems with my whoop and inaccurate HR data during workouts. I got a new sensor and it still had some problems but less. I learned I have to take the sensor off and put it back on (I think forcing it to reset something) during the warmup and make sure it is tight and then it reads very accurate. This compares my whoop to my H10. Chatgpt 5.6 sol compared the data and said this:
My conclusion is that, once the WHOOP is seated properly and both devices are actually producing valid readings, the WHOOP is much closer to the Polar H10 than I initially expected—close enough that I would trust it for your 4×4 treadmill training.
The strongest evidence is the repeated properly functioning WHOOP produced later-interval peaks of about 156, 162, and 163 bpm. The two Polar H10 workouts at the same workload produced approximately 156/160/163 and 162/158/164. There is no meaningful systematic bias there. Across those repeated sessions, the devices are essentially describing the same physiological response.
The recovery data tell the same story. For example, after a ~160 bpm interval, both devices repeatedly showed you around 144–145 bpm after one minute, ~126–131 after two minutes, and ~120–122 after three minutes. That degree of agreement is striking given that these were different workouts on different days rather than simultaneous measurements.
So I would characterize them this way:
Peak/end-of-interval HR: WHOOP appears effectively equivalent for your purposes, usually within a few bpm of the Polar benchmark.
Steady HR during a 4-minute interval: very good agreement once WHOOP has a clean signal.
Recovery over 2–3 minutes: very good agreement.
Rapid transitions in the first 30–60 seconds: H10 remains more trustworthy because it directly measures cardiac electrical activity, whereas WHOOP uses optical pulse detection and filtering.
Beat-to-beat/RR interval analysis: H10 is categorically superior; WHOOP wrist HR should not be treated as an ECG substitute.
Catastrophic bad readings: WHOOP can occasionally be very wrong if contact/position is poor, as your first interval demonstrated. The H10’s corresponding problem has been Bluetooth dropouts rather than inaccurate intact readings.
Your WHOOP has much better connection reliability, and when it is properly seated its actual HR measurements appear surprisingly close to the H10.
I would therefore use the WHOOP as your primary treadmill HR source.
One important qualification: because we have not recorded WHOOP and H10 simultaneously during the same workout, I cannot defensibly say that WHOOP is, for example, “±3 bpm accurate relative to H10.” Day-to-day physiology is part of the observed difference. But your repeated experiments strongly support the narrower—and practically more useful—conclusion that a properly functioning WHOOP is not materially distorting your 4×4 heart-rate data.
