Sleep trackers are good at telling when you are asleep, weaker at telling when you are lying awake, and least reliable at naming sleep stages. In lab studies against polysomnography, the Apple Watch, Oura Ring, Whoop and Fitbit all estimated total sleep fairly well in healthy adults, but deep sleep could be out by 43 minutes.
How does a sleep tracker measure sleep?
Sleep stages are defined by brain waves, eye movements and muscle tone, which is what a full sleep study called polysomnography records (AASM, 2017). A watch or ring cannot see any of those. It infers sleep from signals it can pick up through the skin, such as movement and heart rate, and then estimates which stage you were probably in.
That works well for the broad question of whether you are asleep. It is harder for the finer question of whether you are in light, deep or dreaming sleep, because those stages can look alike from the wrist.
What do the studies show for each device?
Four lab studies compared popular trackers with polysomnography on the same nights.
- A 2024 study of 35 healthy adults tested the Oura Ring Gen3, Fitbit Sense 2 and Apple Watch Series 8. All three detected sleep at least 95% of the time. The Oura Ring did not differ from polysomnography on any sleep stage. The Fitbit underestimated deep sleep by 15 minutes and the Apple Watch by 43 minutes (Robbins, 2024). The lead author sits on Oura's medical advisory board, which the paper declares.
- A 2022 study of 53 adults tested six devices including the Apple Watch S6, Oura Ring Gen 2 and Whoop 3.0. Agreement on sleep or wake was 86% to 89%. Agreement on the exact stage was 53% for the Apple Watch, 60% for Whoop and 61% for Oura, and the authors concluded that all six needed to improve on sleep stages (Miller, 2022).
- A 2023 study of 75 people from a hospital and a sleep clinic tested 11 trackers. Accuracy on sleep stages varied widely between devices, and the Google Pixel Watch and Fitbit Sense 2 performed best on deep sleep (Lee, 2023).
- A 2021 study of 34 healthy young adults tested seven devices, including a Fitbit. Detecting sleep was reliable, detecting wake was poor to moderate, stage estimates were inconsistent, and devices tended to do worse on nights of disrupted sleep (Chinoy, 2021).
Where do trackers go wrong?
The weak point is lying still while awake. In the 2021 study, devices correctly identified only 18% to 54% of the time a person was awake (Chinoy, 2021). A tracker that reads quiet wakefulness as light sleep will overstate how long you slept, and it is likely to do so most on the nights you sleep worst.
Most of these studies were also done in healthy young or middle-aged adults sleeping one to three nights in a lab. Results may differ in people with insomnia, sleep apnoea or other conditions, who were not the focus of this research.
Can a tracker diagnose a sleep problem?
In its 2018 position statement, the American Academy of Sleep Medicine said that consumer sleep technologies "cannot be utilized for the diagnosis and/or treatment of sleep disorders at this time" (AASM, 2018). It added that the data can still be useful when brought to a proper clinical assessment.
Since then, a small number of watch features have been cleared by the US Food and Drug Administration to flag possible sleep apnoea. They look for signs, not a diagnosis, and we explain how well they work in can a smartwatch detect sleep apnoea.
How should I use my sleep data?
Look at trends across weeks, not single nights. A run of short nights, a bedtime that keeps drifting later or a resting heart rate that climbs over a fortnight is more meaningful than one poor score.
Be wary of chasing a number. Sleep doctors have described people seeking treatment for problems they inferred from tracker data, a pattern they named orthosomnia (Baron, 2017). Worrying about a deep sleep figure can make sleep worse. If you want to know what a normal amount looks like, read how much deep sleep do you need.
Keep a consistent wake-up time. It is one of the instructions sleep specialists use in treating insomnia (AASM, 2021), and a tracker makes it easy to check you are doing it.
Bring the data with you if you see a doctor. Together with your history, it can help decide whether a test is worth doing.
When is a sleep study worth it?
A tracker cannot confirm or rule out a sleep disorder. If you snore loudly, someone has seen you stop breathing, you wake unrefreshed most mornings or you are sleepy in the day, a sleep study gives a measured answer. Our guide to the signs of sleep apnoea sets out what to look for.
Seeing a sleep doctor at The Wellness
Our Sleep Clinic in Marylebone starts with a Sleep Consultation for £450, a 60-minute consultation with a doctor, ending with a clear answer on whether you need a sleep study and which one. If a test is needed, the Home Sleep Study at £1,395 is a one-night respiratory study in your own bed, reported by a consultant in respiratory and sleep medicine, with a results consultation and your wearable data reviewed alongside it. The Sleep Clinic programme, chosen by most of the people we see, is £2,995 and brings together the home study, a consultation with a respiratory sleep consultant, blood tests chosen for your sleep and a treatment plan reviewed at three months.
Weighed against that is the cost of carrying on without an answer. Untreated severe sleep apnoea was linked with a higher rate of heart attack and stroke in a ten-year study of men (Marin, 2005), and insufficient sleep costs the UK an estimated 1.36% to 1.86% of its economic output each year (RAND Europe, 2016). Book a sleep consultation when you are ready.
This article is general information and does not replace a consultation. If you are falling asleep at the wheel, stop driving until you have been assessed. If you have chest pain, severe breathlessness or signs of a stroke, call 999.
Frequently asked questions
Which sleep tracker is the most accurate?
No device came out best in every study. In a 2024 study the Oura Ring agreed most closely with polysomnography on sleep stages, while a larger 2023 study found the Google Pixel Watch and Fitbit Sense 2 strongest on deep sleep. All of the main devices estimate total sleep reasonably well in healthy adults.
Is Oura more accurate than Apple Watch for sleep?
In two lab studies the Oura Ring matched polysomnography on sleep stages more closely than the Apple Watch, which underestimated deep sleep by 43 minutes in one of them. The lead author of that study advises Oura, and both studies used small groups of healthy adults, so treat the gap as a guide rather than a verdict.
How accurate is Whoop at tracking sleep?
In a 2022 study of 53 adults, the Whoop 3.0 agreed with polysomnography on whether a period was sleep or wake 86% of the time, and on the exact sleep stage 60% of the time. That is similar to the other devices tested.
Can a sleep tracker tell me if I have a sleep disorder?
Not on its own. The American Academy of Sleep Medicine says consumer sleep technology cannot be used to diagnose or treat sleep disorders. A tracker can show a pattern worth discussing, and a doctor decides whether you need a sleep study.
My tracker says I sleep badly but I feel fine. Should I worry?
How you feel in the day matters more than a score. Sleep doctors have described people seeking treatment for problems they inferred from tracker data alone. If you are alert and well, a low score on its own is not a reason for concern.
Speak to the team about sleep
See what we offer, then book when it suits you.