You strap on your smartwatch, fall asleep, and wake up to a colorful chart showing exactly how much deep sleep, REM sleep, and light sleep you got. It feels authoritative. It feels precise. And millions of people are making decisions about their health based on these numbers.

But should they be?

I spend a lot of time with patients who walk into the sleep clinic carrying screenshots from their Fitbit, Apple Watch, or Oura Ring, convinced they have a sleep disorder — or equally convinced they don’t. It’s created an interesting dynamic, and it’s worth being honest about what these devices can and can’t do.

What Sleep Trackers Actually Measure

First, let’s clarify what’s happening under the hood. Consumer wearables don’t measure brain activity. Polysomnography — the clinical gold standard — uses electroencephalography (EEG), electromyography (EMG), and electrooculography (EOG) to directly classify sleep stages based on electrical signals from the brain, muscles, and eyes.

Your wristwatch measures movement (accelerometry) and, in newer models, heart rate and heart rate variability. Some devices add blood oxygen saturation (SpO2) or skin temperature. Algorithms then infer sleep stages from these proxy signals.

That distinction — measuring versus inferring — is crucial.

The Validation Studies

There’s actually a decent body of research comparing consumer trackers to polysomnography, and the findings are fairly consistent across devices and studies.

Total sleep time: Pretty good. Most modern trackers estimate total sleep time within 20-30 minutes of polysomnography values, on average. That’s reasonable for a consumer product. They’re especially good at detecting when you’re asleep versus awake, at least during the middle of the night when you’re sleeping soundly.

Sleep onset: Decent but not great. Trackers tend to overestimate how quickly you fall asleep. If you’re lying still in bed but not yet sleeping, the accelerometer sees stillness and often logs it as sleep. This matters because sleep onset latency is a clinically important metric.

Wake after sleep onset (WASO): Underestimated. This is where trackers struggle most. Brief awakenings — the kind that fragment sleep in conditions like sleep apnea — are frequently missed. If you wake up for 30 seconds, roll over, and fall back asleep, the device may never register that you were awake. A validation study published in the journal Sleep found that most consumer devices significantly underestimate WASO, which means they paint a rosier picture of sleep quality than reality warrants.

Sleep staging: Take it with a grain of salt. This is where I’d urge the most caution. The accuracy of deep sleep and REM classification varies enormously between devices and between nights. A comprehensive review from the National Sleep Foundation noted that while some devices achieved moderate agreement with polysomnography for REM detection, deep sleep classification was inconsistent and often unreliable at the individual level.

On any given night, your tracker might say you got 90 minutes of deep sleep when polysomnography would show 45. Or vice versa. The numbers bounce around enough that drawing clinical conclusions from them is premature.

Where Trackers Are Genuinely Useful

I’m not trying to dismiss consumer sleep technology entirely. There are real benefits:

Trend tracking over weeks and months. Even if the absolute numbers are off, the relative trends can be informative. If your average total sleep time drops from 7 hours to 5.5 hours over a month, that’s a meaningful signal regardless of whether the exact minutes are perfectly accurate.

Motivating behavior change. Many people pay more attention to their sleep habits after getting a tracker. If seeing your data makes you prioritize consistent bedtimes or cut back on evening caffeine, the device has served a useful purpose.

SpO2 screening. Some newer devices track overnight oxygen saturation and can flag patterns suggestive of sleep-disordered breathing. This isn’t diagnostic, but it can prompt someone to get a formal evaluation who might otherwise not have. Apple Watch’s FDA-cleared sleep apnea notification feature is a notable step in this direction.

Sleep diary replacement. For patients undergoing cognitive behavioral therapy for insomnia, a tracker can supplement (not replace) subjective sleep diaries, especially for people who find manual tracking burdensome.

Where They Cause Problems

The flip side is real too. I’ve seen patients become anxious about their sleep data — a phenomenon researchers have dubbed “orthosomnia.” They fixate on optimizing their deep sleep percentages, become stressed when the numbers don’t look right, and actually sleep worse because of the monitoring itself.

I’ve also had patients refuse sleep testing because “my watch says I’m fine,” when their clinical symptoms clearly warranted evaluation. A device that misses brief awakenings and underestimates fragmentation can falsely reassure someone who actually has moderate sleep apnea.

My Advice

Use your sleep tracker for what it’s good at — noticing trends, building awareness, and staying motivated. Don’t treat the sleep stage breakdown as medical data. And if you’re experiencing symptoms like excessive daytime sleepiness, loud snoring, witnessed apneas, or unrefreshing sleep, no consumer device is a substitute for a proper clinical evaluation.

The technology is getting better every year. But we’re not at the point where a wristwatch replaces a sleep study. Not yet.