What is a good HRV? Why Apple Watch, WHOOP, Oura and Garmin show different numbers

Your HRV this morning is 38. Your friend’s is 92. You both slept eight hours, you both trained yesterday, and your Apple Watch and their WHOOP seem to agree that their nervous system is in far better shape than yours.

They do not agree on anything. Those two numbers were calculated with different maths, at different times, by devices that were never designed to be compared. Here is what HRV actually measures, why every device reports it differently, and what counts as good for you.

What HRV is measuring

Your heart does not tick like a metronome. Even at a steady 60 beats per minute, the gap between beats stretches and shrinks by tens of milliseconds, and heart rate variability is the size of that wobble. More wobble generally means your rest-and-digest system has the upper hand. Less tends to show up after short sleep, alcohol, stress, illness or a heavy training block.

The range is enormous. Oura puts typical values anywhere from under 20 to over 200 milliseconds, and its average across members in 2026 was about 41 ms, 42.8 for men and 39.1 for women. HRV also declines with age, which Garmin builds into its own status labels. That is why “what is a good HRV” has no single answer: a number that is excellent at 55 can be ordinary at 25.

Two kinds of maths: RMSSD and SDNN

Before devices even disagree about timing, they disagree about the formula.

• RMSSD compares each beat interval with the one right before it. It is the measure researchers use for the vagal, rest-and-digest side of your nervous system, and it is what WHOOP, Oura and Fitbit say they use. Garmin’s HRV explainer describes RMSSD too.
• SDNN measures the overall spread of every beat interval in a window. Both halves of the nervous system feed into it, and it grows with the length of the recording. Apple Health reports HRV as SDNN.

These are not two names for one number. A widely cited review of HRV metrics warns that it is inappropriate to compare SDNN values taken over windows of different length, and that 24-hour, short and ultra-short norms are not interchangeable. Even in the lab, the pooled norms for a five-minute recording come out at about 50 ms for SDNN and 42 ms for RMSSD, so the two sit on different scales before a wearable is involved at all.

When each device takes the reading

Read the Apple row twice, because it is changing. For years the watch sampled HRV in short readings while you were still, by default every four hours, plus any time you opened a Breathe session. Apple’s September 2026 update says Series 12 and Ultra 4 now measure HRV as often as every five minutes, and add a Recovery HRV that is analysed overnight against your own baseline. Apple has not said which statistic Recovery HRV uses. Until it does, treat an Apple number and an Oura number as two different instruments, not one instrument read twice.

Timing matters as much as the formula. A reading snatched at your desk after a coffee and an average across a full night of sleep are measuring your nervous system in two different states. Neither is wrong. They just answer different questions.

How to read your own HRV

The good news is that every device converges on the same advice, because every one of them reads HRV against you.

• Nightly HRV
• Seven-day average
• Your baseline

Illustrative shapes, not measured data. Fourteen nights.

• Give it a baseline first. Garmin’s HRV Status needs about three weeks of consistent sleep data, Samsung builds its baseline over seven nights, and WHOOP compares you with the last 30 days. A number with no baseline behind it has nothing to be read against.
• Ignore single nights. Look at the dip on day five in the chart above: one bad night, already recovered. The run of low nights in week two is the thing worth noticing, which is why Garmin judges your seven-day average and Oura weights your last 14 days.
• Act on runs, not readings. Several days below your baseline usually means sleep, stress, alcohol, illness or training load caught up with you. An easier day and an earlier night is the standard response.
• Do not switch devices mid-trend. Moving from an Apple Watch to a ring swaps SDNN for RMSSD and daytime readings for overnight ones. Start a fresh baseline rather than reading the new numbers against the old.

HRV belongs with the other overnight numbers that only make sense against yourself, resting heart rate and your sleep score, and it is the main ingredient in every branded recovery and readiness score, which is why those do not compare between people either.

Fit, the cross-brand fitness competition app, treats HRV exactly this way, and since it is our app, weigh this accordingly. Fit reads HRV from whichever tracker you wear and shows today’s number against your own normal. On a crew’s board it appears only as health context, never a podium, and its note flags the catch from this post: Apple reports SDNN while WHOOP, Garmin, Oura and Fitbit report RMSSD, so a gap between two devices can be the maths rather than the heart. The crew races on effort instead, using the units that carry between brands.

What to do with this

1. Stop comparing your HRV with anyone else’s. Age and genetics spread healthy people from 20 to over 200 milliseconds, and devices do not even use the same formula.
2. Wear one device the same way, every night. Overnight averages are steadier than spot readings, and consistency is what makes a baseline.
3. Read the weekly direction. Above or at baseline, carry on. A run of days below it, back off and sleep.
4. Pair it with resting heart rate. HRV falling while resting heart rate climbs is one of the clearest recovery signals a wearable can give you, on any brand.