Your smartwatch tells you that last night you got 6 hours and 51 minutes of sleep, including 58 minutes of deep sleep and 84 minutes of REM. It feels precise. It feels medical. But here is the honest answer to whether you should trust those numbers: partly.
A modern sleep tracker is genuinely good at one thing — telling whether you were asleep or awake — and much weaker at the thing people care about most: breaking your night into deep, light, and REM stages. Understanding that gap is the difference between a tracker that helps you build better habits and one that just makes you anxious about a bad "sleep score."
This article is general information, not medical advice. If you have ongoing sleep problems — loud snoring, gasping awake, or persistent daytime exhaustion — talk to a qualified healthcare professional.
How a Wearable Even Knows You're Asleep
A clinical sleep study measures sleep directly. A wristband or ring measures it indirectly — and that distinction explains almost everything about where trackers succeed and where they fail.
Consumer wearables lean on two main sensors:
- An accelerometer that detects movement. Lie still and the device leans toward "asleep"; toss, turn, or reach for your phone and it leans toward "awake." This motion-based approach is called actigraphy, and sleep researchers have used it for decades.
- A photoplethysmography (PPG) sensor — the green LEDs on the back of the device. It measures your pulse by reading light bounced off the blood under your skin, which yields heart rate and heart-rate variability (HRV). Some rings add skin temperature and respiration signals.
An algorithm then blends motion, heart rate, and HRV to estimate which sleep stage you are in. Your heart rate and movement do genuinely shift between light sleep, deep sleep, and REM — so the guess is informed, not random. But it is still a guess based on the body's downstream signals, not on the sleeping brain itself.
That is the crucial contrast with the medical gold standard, polysomnography (PSG), the overnight test done in sleep labs. PSG places electrodes on the scalp to record brain waves (EEG), plus sensors for eye movement (EOG) and muscle tone (EMG). Sleep stages are defined by brain-wave patterns, so PSG measures the very thing wearables can only infer.
So How Accurate Are They, Really?
Here the guesswork pays off — for the simplest question. When researchers at Brigham and Women's Hospital ran a 2024 study putting three popular devices (an Oura Ring Gen 3, a Fitbit Sense 2, and an Apple Watch Series 8) head-to-head against PSG on 35 healthy adults, every device detected sleep versus wake with a sensitivity of 95% or higher. In plain terms: if you were asleep, the tracker almost always knew it.
The trouble starts when the same devices try to name the stage of sleep. Across those tests, sensitivity for correctly classifying a given stage dropped to a range of roughly 50% to 86%, depending on the device and the stage. Deep sleep and wake — the two stages people fixate on — were the hardest to pin down.
A useful way to judge stage accuracy is Cohen's kappa, a score from 0 (random guessing) to 1 (perfect agreement) that adjusts for lucky guesses. In the Brigham study the four-stage kappas landed around 0.65 for the Oura Ring, 0.60 for the Apple Watch, and 0.55 for the Fitbit — "moderate" agreement with the sleep lab, not the near-perfect match the tidy pie charts in the app imply. A separate 2025 study of six wrist wearables reported kappas ranging from just 0.21 to 0.53, underscoring how much results vary by device and algorithm.
| What the tracker is doing | Typical accuracy vs. sleep lab | Verdict |
|---|---|---|
| Detecting sleep vs. wake | ~95%+ sensitivity | Trustworthy |
| Estimating total sleep time | Within a modest margin | Useful ballpark |
| Classifying the exact stage (deep / light / REM) | ~50–86% sensitivity; kappa ~0.55–0.65 | Directional only |
| Detecting brief wake-ups | Often missed (low wake specificity) | Weakest area |
Why Sleep-vs-Wake Is Easy but Stages Are Hard
Two quirks explain the pattern. First, wearables tend to assume that stillness equals sleep. Lie quietly awake — the classic 2 a.m. ceiling-stare — and a motion-based device often scores it as sleep. That is why "wake specificity" is low and why trackers gently overestimate how much you slept. Second, deep sleep and REM produce distinctive brain-wave signatures but only subtle differences in heart rate and movement, so an algorithm working from the outside has little to grab onto. It is reading the shadow of sleep, not sleep itself.
What Sleep Trackers Are Genuinely Good For
None of this makes them useless — far from it. Used for the right jobs, a tracker is one of the most practical health tools you can wear.
- Spotting trends over weeks, not judging a single night. One night's stage breakdown is noisy; the four-week trend in your total sleep is where the signal lives.
- Measuring consistency. Trackers are excellent at showing whether your bed and wake times wander — and sleep regularity is one of the most powerful, evidence-backed levers for feeling rested. (See our guide on how much sleep you actually need.)
- Estimating total sleep time. The headline "you slept 6h 51m" is far more reliable than the deep-sleep sub-total beneath it.
- Nudging better habits. The simple act of seeing your sleep can push an earlier bedtime or less late-night scrolling — the same behavior-change effect that step counts have on movement, which we cover in desk-job fitness for professionals.
What They Can't Do — and the "Orthosomnia" Trap
The line a good user never crosses is treating a consumer tracker as a diagnostic device. The American Academy of Sleep Medicine is explicit on this point: because these gadgets lack clinical validation and, for sleep-staging, FDA clearance, they "cannot be utilized for the diagnosis and/or treatment of sleep disorders." A tracker cannot rule in or rule out sleep apnea, insomnia, or restless legs. It can only prompt a conversation you then have with a professional.
There is also a subtler risk. Sleep physicians have described a pattern they nicknamed orthosomnia — from ortho ("correct") and somnia ("sleep") — in a 2017 case series of patients who became so fixated on achieving perfect tracker scores that the anxiety itself worsened their sleep. Chasing a number can become the problem it was meant to solve. If your device is making you dread bedtime, that is a signal to loosen your grip on the data, not tighten it.
How to Actually Use Your Sleep Tracker Well
- Watch the trend line, ignore the nightly drama. Look at weekly and monthly averages, not last night's exact figures.
- Prioritize the numbers it gets right: total sleep time and the consistency of your bed/wake times.
- Don't diagnose yourself. Recurring loud snoring, choking or gasping awake, or heavy daytime sleepiness are reasons to see a doctor — regardless of what your "sleep score" says.
- Treat stage percentages as estimates. If deep sleep looks low one night, that may be the algorithm, not your brain.
- Protect your relationship with sleep. If the app is breeding anxiety, turn off the score and keep only the bedtime reminder.
The Bottom Line
Sleep trackers work — for the job they are actually built to do. They are reliable at telling whether you slept and roughly how long, very good at revealing patterns over time, and genuinely helpful at nudging better habits. They are only fair at slicing your night into precise stages, and they are not medical instruments. Use your wearable as a coach, not a diagnostician: follow the trends, act on the habits, and take the deep-sleep minutes with the pinch of salt they deserve.
This is the first piece in our wearable health-tech series. Next we look at a higher-stakes question — can your smartwatch detect a heart problem? — where the same sensors do something genuinely useful, with one dangerous limit.
Frequently Asked Questions
Are sleep trackers accurate?
It depends on what you ask them. For detecting whether you are asleep or awake, modern wearables are highly accurate — around 95% or better against a clinical sleep study. For estimating your total sleep time they are a reliable ballpark. But for classifying the exact sleep stage — deep, light, or REM — accuracy falls to roughly 50–86%, so those specific minutes are best treated as informed estimates rather than facts.
Why does my sleep tracker say I got deep sleep when I felt awake?
Wearables infer sleep mainly from stillness and heart rate, not from brain waves. If you lie quietly in bed without moving, the device often scores that time as sleep even if you were awake — a known weakness called low "wake specificity." It is also why trackers tend to slightly overestimate how much you slept.
How is a wearable different from a real sleep study?
A clinical sleep study (polysomnography) places electrodes on your scalp to record brain waves, which is how sleep stages are actually defined, plus sensors for eye movement and muscle activity. A wearable has none of that — it estimates sleep from motion and pulse. So the sleep lab measures your sleep stages while a wearable guesses them from indirect signals.
Can a sleep tracker diagnose sleep apnea or insomnia?
No. The American Academy of Sleep Medicine states that consumer sleep trackers cannot be used to diagnose or treat sleep disorders because they lack clinical validation. A tracker might hint that something is off — restless nights, frequent waking — but only a qualified professional, usually with a proper sleep study, can diagnose a sleep disorder.
Which sleep tracker is the most accurate?
In a 2024 Brigham and Women's Hospital comparison, a finger-worn Oura Ring edged out the Apple Watch and Fitbit for four-stage sleep classification, but all three landed in "moderate agreement" territory with the sleep lab. The honest takeaway is that no consumer device matches a clinical study, and the gaps between the leading brands are smaller than the gap between any of them and the medical gold standard.
Should I stop using my sleep tracker?
Not necessarily. If it helps you keep a consistent schedule and build better habits, it is doing valuable work. The one thing to avoid is anxiety over the exact scores — a pattern clinicians call orthosomnia, where obsessing over perfect data actually harms sleep. Focus on the trends, and if the app stresses you out, hide the score and keep the bedtime reminder.
Sources
- Accuracy of Three Commercial Wearable Devices for Sleep Tracking in Healthy Adults — Sensors (2024), Brigham and Women's Hospital
- A performance validation of six commercial wrist-worn wearable sleep-tracking devices — SLEEP Advances (2025), Oxford Academic
- Consumer Sleep Technology: An American Academy of Sleep Medicine Position Statement — Journal of Clinical Sleep Medicine
- Orthosomnia: Are Some Patients Taking the Quantified Self Too Far? — Journal of Clinical Sleep Medicine (2017)
- Study finds Oura Ring most accurate in four-stage sleep classification (Sleep Review)
