WHOOP explores respiratory rate variability as a new recovery metric

WHOOP is exploring respiratory rate variability, or RRV, as another way to judge recovery and sleep. A newly published patent describes measuring the changing timing between breaths and using that information alongside other physiological data.

Basically, WHOOP is looking beyond the familiar respiratory rate number. Instead of only counting how many times you breathe per minute, the system looks at how much the length of one respiratory cycle changes compared with the next.

WHOOP wants to measure variability between breaths

The patent, called “Respiratory Cycle Onset Detection”, was published on September 24 and originally filed in March 2025. The title sounds fairly technical, but the more interesting part is what WHOOP wants to do once it identifies those individual breaths.

The system calculates an RRV score based on changes between respiratory cycles. WHOOP describes several possible calculations, including standard deviation, successive differences, RMSSD, coefficient of variation and median absolute deviation.

There is a loose parallel with HRV here. HRV measures changes in the time between heartbeats, while RRV measures changes between respiratory cycles. They are different physiological signals, but both try to extract more information from variability rather than simply looking at an average rate.

That fits with the way WHOOP already handles recovery. The company combines several measurements rather than relying on a single sensor reading, which is one reason it still stands apart from some of the newer devices in our best screenless fitness tracker guide.

Recovery appears to be a major use case

The patent directly describes using RRV to determine a recovery state. In one example, lower respiratory variability can correspond to greater recovery, while higher variability can indicate lower recovery.

WHOOP also describes combining RRV with measurements such as HRV, sleep scores and recent strain. That suggests RRV could eventually become another input into the existing Recovery calculation rather than appearing as a completely separate headline metric.

The system can also compare current RRV against previous readings and look for longer-term trends. That would allow WHOOP to track whether respiratory patterns are shifting across days or weeks rather than judging a single night in isolation.

Sleep is another potential use. The patent describes using respiratory rate variability to help determine sleep stages, either through thresholds or a machine-learning model.

Existing sensors may be enough

Perhaps the most practical part is that this does not appear to require a new sensor. WHOOP describes extracting respiratory information from changes in the pulse waveform recorded by a wearable physiological monitor.

The patent also describes using accelerometer data to detect breathing-related movement when the wearer is relatively still. Sleep is an obvious opportunity because movement is limited and the device has a long window of continuous data to analyse.

So this could potentially be a software feature built from signals WHOOP hardware already collects. Perhaps via a firmware update. Definitely something to keep an eye on.