Apple Watch tracks sleep using a combination of its built-in accelerometer, heart rate sensor, and machine learning algorithms that work together to determine when you’re asleep, how long you sleep, and what sleep stage you’re in. It does all of this passively from your wrist, requiring only that you wear it to bed with at least 30% battery and have sleep tracking enabled.
Motion and Heart Rate: The Core Sensors
The accelerometer is the foundation of Apple Watch sleep tracking. This motion sensor detects the tiny movements of your wrist throughout the night, distinguishing between the stillness of sleep and the restlessness of being awake. When you stop moving in consistent patterns, the watch begins classifying that time as sleep.
Heart rate adds a second layer of data. During sleep, your heart rate drops and follows predictable patterns that differ between sleep stages. During deep sleep, your heart rate is at its lowest and most steady. During REM sleep (when dreaming occurs), it becomes more variable, closer to waking levels. The optical heart rate sensor on the back of the watch pulses green LED light into your skin and measures how much light is absorbed by blood flowing through your wrist, calculating your pulse continuously through the night.
Apple’s algorithms combine these motion and heart rate signals to classify your sleep into three stages: REM, Core (a combination of the lighter sleep stages), and Deep. The watch also identifies periods of wakefulness, so your sleep timeline shows interruptions if you got up or stirred significantly during the night.
How Accurate Is Sleep Stage Detection?
A 2025 validation study published in SLEEP Advances compared six commercial wearables, including the Apple Watch Series 8, against polysomnography, the gold-standard sleep test performed in a lab with electrodes attached to your scalp. The Apple Watch correctly identified sleep epochs 96.27% of the time, meaning it’s very good at knowing when you’re actually asleep.
Where it struggles is detecting when you’re awake. The watch correctly classified only about 52% of wake periods recorded by the lab equipment. That was actually the best performance among the six devices tested, but it means the Apple Watch tends to overcount your sleep. If you’re lying still in bed but awake, the watch will often mark that time as sleep. For most people, the overall sleep duration the watch reports will be slightly inflated compared to how much sleep you truly got.
Wrist Temperature Tracking
Apple Watch Series 8 and later models include a temperature sensor that takes readings from your wrist overnight. This isn’t used to classify sleep stages directly, but it adds a health metric to your sleep data. The watch needs about five nights of data to establish your personal baseline, and it requires at least four hours of tracked sleep each night to collect a usable reading.
Once your baseline is set, the watch shows nightly deviations in the Vitals app, so you can spot when your temperature is outside your typical range. Elevated wrist temperature can signal illness, stress, or alcohol consumption. For people who track their menstrual cycle, the watch uses these temperature shifts to improve period predictions and provide ovulation estimates, since body temperature rises slightly after ovulation.
Breathing Disturbances and Sleep Apnea
Starting with watchOS 11, newer Apple Watch models can monitor breathing disturbances during sleep. The motion sensors measure subtle chest and wrist movements associated with breathing, calculating your breaths per minute. The watch logs how often your breathing is disrupted, and over time it categorizes your breathing disturbance levels as elevated or not.
Some models also include a blood oxygen sensor that uses pulse oximetry, shining red and infrared light into your wrist to estimate blood oxygen levels. However, it only takes occasional measurements throughout the night and requires your arm to be in the correct position, so it doesn’t provide continuous monitoring.
Apple received FDA clearance for the Breathing Disturbances feature in September 2024, but that clearance is specifically for early detection, not diagnosis. The watch can flag signs of possible sleep apnea and suggest you talk to a doctor, but it cannot diagnose sleep apnea on its own. A formal diagnosis still requires a clinical sleep study.
What You Need for It to Work
Sleep tracking on Apple Watch isn’t fully automatic out of the box. You need to set up a sleep schedule in the Health app on your iPhone, which activates Sleep Focus mode at bedtime. The watch needs to be charged to at least 30% before bed, which means most people develop a habit of charging during their morning routine or in the evening before sleep.
The watch tracks sleep whenever Sleep Focus is active and you’re wearing it. In the morning, you’ll see a summary showing your total time asleep, time in each sleep stage, and your respiratory rate. Over weeks and months, trends become visible in the Health app, letting you spot patterns like consistently short deep sleep or frequent overnight wake periods. The Vitals app on newer models pulls together heart rate, respiratory rate, wrist temperature, and blood oxygen into a single overnight summary, flagging any metric that falls outside your personal norm.
What the Watch Can and Can’t Tell You
Apple Watch gives you a reasonable picture of your sleep duration and a rough breakdown of sleep stages. It’s useful for spotting trends: Are you getting more or less sleep this month? Is your deep sleep consistently low? Are your breathing disturbances increasing? These patterns can prompt meaningful conversations with a doctor or motivate changes to your sleep habits.
What it can’t do is match the precision of a clinical sleep study. It tends to overestimate sleep time, its sleep stage classifications are approximations rather than exact measurements, and its blood oxygen readings are intermittent. Think of it as a reliable night-to-night tracker for general patterns rather than a medical-grade sleep lab on your wrist.