Sleep efficiency: the number that matters more than hours

Sleep efficiency is total sleep time divided by total time in bed, expressed as a percentage. If you are in bed for 8 hours and asleep for 7, your efficiency is 88%. It is the single number that separates 'time asleep' from 'time lying there', which is why sleep clinicians and trackers report it: two people can both spend 8 hours in bed and get very different amounts of actual sleep. Time in bed starts when you lie down intending to sleep and ends when you get up for the day, so reading, scrolling, or lying awake at 3am all count against the number.

Roughly 85% is the conventional clinical threshold for healthy sleep, and good sleepers commonly sit around 90 to 95%. Below about 85% points to fragmented sleep or too much time awake in bed, the pattern seen in insomnia. The number drifts down with age as sleep naturally fragments, so an older adult at 82% is not necessarily disordered. Treat 85% as a floor to aim above, not a target to obsess over on any single night, because night-to-night variability is large and normal.

Efficiency says nothing about how much you slept, only what fraction of your time in bed was sleep. Efficiency of 95% on 5 hours in bed is still 4.75 hours of sleep, which for most adults is chronic restriction (see the sleep-debt guide for what that costs). In fact, a very high efficiency, above roughly 95%, often signals sleep deprivation: when you are sleep-deprived you fall asleep almost instantly and barely wake, which inflates the percentage. Always read efficiency alongside total sleep time, never on its own.

Consumer wearables like Oura, Whoop, and the Apple Watch estimate sleep efficiency from movement and heart-rate patterns rather than EEG, so they infer when you were awake in bed rather than measuring it. Against polysomnography they tend to overestimate efficiency, because brief quiet wakefulness reads as sleep to a motion-and-heart-rate model. The number is still useful as a personal trend: watch whether your own efficiency is climbing or falling over weeks, rather than trusting a single night's figure or comparing your device to someone else's. For how these devices stack up, see the sleep-tracker comparison.

The most effective way to raise a low efficiency is counterintuitive: spend less time in bed, not more. Bedtime compression, the core of sleep restriction therapy in CBT-I (Spielman et al. 1987), temporarily shrinks your time in bed to match the sleep you are actually getting. That concentrates sleep pressure, so you fall asleep faster and wake less, which pushes efficiency up. Once efficiency holds above 85 to 90% for about a week, you widen the window in 15-minute steps. It feels harsh for the first few nights, but it is one of the best-evidenced behavioral tools for consolidating broken sleep.

Bedtime compression works best alongside the basics: a fixed wake time seven days a week, a dose of morning light to anchor the clock (see the morning-sunlight protocol), a caffeine cutoff timed to its long half-life (covered in the caffeine-and-sleep guide), and getting out of bed when you cannot sleep rather than lying there watching the clock, which is the stimulus-control rule that protects the bed-equals-sleep association. The reliable efficiency killers are alcohol (it fragments the second half of the night), long or late naps, and using the bed for wakeful activities. If you keep waking in the small hours despite this, see why-you-wake-at-3am.