Best tracker for HRV: how each wearable measures it
Heart-rate variability is the reading people buy a tracker for and then struggle to use. Part of the trouble is that every device measures it differently, so the number on one is not the number on another. This guide explains how each one does it and which to choose if HRV is the point.
Your heart does not beat like a metronome. The gap between beats changes slightly from one beat to the next, and the size of that change is heart-rate variability. A larger variation generally goes with a body that is rested and able to adapt; a smaller one with strain, illness, alcohol or poor sleep. It is personal: a value that is low for one person is normal for another, so the only useful comparison is with your own baseline.
It is personal. The only useful comparison is with your own baseline.
| Device | When it measures | The statistic | Shown as |
|---|---|---|---|
| Oura ring | All night, while asleep | RMSSD, averaged over the night | Average HRV, plus a night curve and a long-term trend |
| WHOOP | All night, weighted to the last deep-sleep period | RMSSD | A single overnight value feeding Recovery |
| Apple Watch | Sporadic samples through the day and night, plus during Mindfulness sessions | SDNN, a different statistic | A list of samples in the Health app; not an overnight average |
| Fitbit | Overnight, during deep sleep, on recent models | RMSSD | A nightly value in Health Metrics |
| Polar | The first hours of sleep, as part of Nightly Recharge | RMSSD-based | ANS charge and an HRV value on supported watches |
| Suunto | Overnight, on recent watches | RMSSD | A nightly value and a range against your recent norm |
| Withings | Limited; not a focus | — | Not a reason to buy Withings |
| Garmin | All night, five-minute averages, on recent watches | RMSSD | HRV Status, with a personal baseline range. Not connected to Humotron yet |
Two things follow. First, RMSSD and SDNN are different statistics, so an Apple Watch HRV cannot be compared with an Oura or WHOOP HRV at all; the numbers are not measuring the same thing. Second, even among the RMSSD devices, the window differs: an all-night average (Oura) runs steadier than a last-deep-sleep sample (WHOOP), which reacts faster to yesterday.
Against a chest strap or an ECG, overnight RMSSD from a finger or wrist sensor is close enough to be useful when the body is still, which is why every maker measures it during sleep. Daytime readings from a wrist are much noisier. Published comparisons put Oura, WHOOP and recent Fitbit and Garmin devices within a usable margin of a reference overnight; Apple Watch’s sporadic SDNN samples are harder to use as a trend because the timing varies.
Any device will show you a lower number after alcohol, a late meal, a hard session or the start of a cold. That is the reading working, not failing.
Whatever you choose, give it a month before you read anything into it. The baseline is the product; the first fortnight is just the device learning what your normal looks like.
Humotron reads your overnight variability as one of four daily summaries, alongside sleep, resting heart rate and activity, over the last six months. It never compares your HRV with a population. It looks for your own pattern: a fall over three weeks with training steady, a swing that tracks your bedtime, a value that has drifted and stayed there. Each match says which of those it was built on and which reading should move if the match works.
If your HRV is already low and you want to know what to try, that is its own guide: Low HRV: what your data says and what to try.
None of this is medical advice. A resting heart rate or HRV that stays out of its usual range for more than two or three weeks, a blood-oxygen reading that is repeatedly low, or any reading that comes with symptoms, is a conversation for a GP, not a tracker.
What is a good HRV score?
There is no good number. Values range from under 20 to over 100 milliseconds across healthy people and fall with age. The useful question is whether yours is in its own usual range.
Why is my Apple Watch HRV so different from my Oura?
They measure different statistics, SDNN and RMSSD, at different times of day. They are not comparable and never will be.
Is a ring more accurate for HRV than a watch?
For overnight readings they are close; the finger gives a slightly steadier signal and the ring is less likely to be loose. For daytime readings neither is reliable.
When should HRV be measured?
While asleep, which is when every device above does it. A morning reading taken sitting still is the older method and still works; a reading during the day is mostly noise.