Sleep efficiency is the share of your time in bed that you actually spend asleep, written as a percentage. If you were in bed for eight hours and asleep for seven, your sleep efficiency was about 87 percent. It is one of the few sleep numbers that has a real clinical history behind it rather than a proprietary algorithm — which makes it more useful than most of what your wearable shows you, and also easier to misread.
A National Sleep Foundation expert panel that reviewed 277 studies concluded that a sleep efficiency of 85 percent or higher is generally an appropriate indicator of good sleep quality across every age group, while efficiency below 75 percent points toward poor sleep quality (Ohayon et al., 2017, Sleep Health). That threshold is worth knowing. It is also worth knowing why it might not apply to your life right now.
How is sleep efficiency calculated?
The formula is simple:
Sleep efficiency = (total sleep time ÷ time in bed) × 100
So seven hours of sleep across eight hours in bed gives you 87.5 percent. Six hours of sleep across nine hours in bed gives you 67 percent — the same amount of sleep, a very different number, because the denominator changed.
That denominator is where things get slippery. Taken literally, “time in bed” includes everything you do in bed: reading a chapter, answering messages, talking with a partner, lying awake at 5 a.m. deciding whether to get up. Researchers have pointed out that this makes the standard formula inconsistent between studies and between people, and have proposed measuring against the time you actually intended to be asleep instead (Reed & Sacco, 2016, Journal of Clinical Sleep Medicine).
This matters practically. If you read for forty minutes in bed every night, your sleep efficiency will look worse than someone who reads on the sofa and gets identical sleep. The number dropped; nothing about your sleep did.
What counts as a good sleep efficiency?
The 85 percent benchmark is the most widely cited figure, and it holds up as a population-level marker. But two things complicate it.
Age moves the baseline. Sleep becomes lighter and more fragmented across adulthood in ways that have nothing to do with doing anything wrong. A meta-analysis of 65 studies covering 3,577 healthy people from childhood to old age found that wake time after sleep onset showed the largest age-related change of any sleep parameter — a steady increase of roughly ten minutes per decade between ages 30 and 60 (Ohayon et al., 2004, Sleep). Ten extra minutes of nighttime wakefulness per decade, with time in bed unchanged, means efficiency drifts down on its own. That drift is normal biology, not decline you need to reverse. We go deeper into this in why sleep gets worse with age.
The number can be high and still unhelpful. Someone sleeping five hours in five and a half hours in bed has 91 percent efficiency and is still short on sleep. Efficiency describes the quality of the fit between bed and sleep. It says nothing about whether you got enough, which is a separate question covered in how much sleep do you need.
Why is my sleep efficiency low?
Low efficiency almost always comes from one of three places, and they call for different responses.
Time to fall asleep. If it routinely takes you more than half an hour to drop off, that time is counted as awake in the denominator. Racing thoughts, a too-early bedtime, and a body clock set later than your schedule are the usual culprits — how to fall asleep faster covers the levers that actually move this one.
Wake time during the night. Every awakening you have, whether you remember it or not, comes out of the numerator. Hot flashes, pain, reflux, a baby, a full bladder, and a partner who snores all land here.
Time in bed that was never going to be sleep. This is the most commonly overlooked cause and the easiest to fix. Extra hours in bed hoping sleep will come, long lie-ins on weekends, scrolling for forty minutes before lights-out — all of it inflates the denominator without adding sleep. The irony is that the instinct to spend more time in bed when you are sleeping badly is the single most reliable way to make efficiency worse.
How does sleep efficiency work for new parents?
This is where the 85 percent benchmark most deserves a caveat, because the research on early parenthood shows something genuinely reassuring.
In a study that tracked mothers with wrist actigraphy across the first sixteen postpartum weeks, average nocturnal sleep time held steady at about 7.2 hours and did not change significantly over that period. What changed was fragmentation. Sleep efficiency was 79.7 percent at postpartum week two, and had risen to 90.2 percent by week sixteen, tracking a fall in sleep fragmentation over the same window (Montgomery-Downs et al., 2010, American Journal of Obstetrics & Gynecology).
Two things are worth pulling out of that. First: at two weeks postpartum, a sleep efficiency well under the “good sleep” threshold was the normal finding, not a warning sign. Second: the total amount of sleep barely moved while the experience of it changed enormously. The exhaustion of the newborn weeks is not mainly a shortage of hours — it is the shattering of those hours into pieces, which is exactly what efficiency captures and what a duration number misses. That is the same mechanism behind sleep debt for parents.
If you are in the first months with a baby, a low efficiency number is describing your circumstances accurately. It is not a report card. The most useful thing it can tell you is whether the fragmentation is slowly easing, which the research suggests it does.
One practical note: if you get up to feed and then stay in bed awake for another forty minutes, that counts against you twice. Getting out of bed once settled — or accepting the night as it is rather than lying there waiting — often does more for the number than any sleep hygiene change. How to get back to sleep has the specifics.
How does perimenopause affect sleep efficiency?
Through the same mechanism, from a different cause. During the menopausal transition, sleep difficulties are linked to menopausal stage and to changes in follicle-stimulating hormone and estradiol over and above the effects of age alone — and hot flashes are a major contributor, with many of them linked to awakenings visible on polysomnography. Hot-flash-associated wake time contributes significantly to overall wake after sleep onset (Baker et al., 2018, Nature and Science of Sleep). Roughly 26 percent of women in this transition report symptoms severe enough to meet criteria for insomnia.
The pattern this produces is specific: normal or near-normal time in bed, plenty of intent to sleep, and a scattering of brief awakenings that drag the numerator down. You can be in bed nine hours and register a poor efficiency number, which is one of the more demoralizing experiences a tracker can hand you.
There is also an age effect stacked underneath the hormonal one — the ten-minutes-per-decade drift in nighttime wakefulness applies here too. So a woman at 48 comparing her number to her own at 32 is comparing across two changes at once. Perimenopause and sleep and night sweats and sleep cover what actually helps with the awakenings themselves.
How accurate is the sleep efficiency on my tracker?
Cautiously. Consumer wearables do reasonably well on total sleep time but have a known blind spot exactly where efficiency is decided: they systematically miss brief awakenings, because their algorithms read stillness plus a low heart rate as sleep (Chinoy et al., 2021, Sleep). If you wake at 3 a.m. and lie still rather than moving around, there is a good chance the device scores you as asleep.
The direction of that error is worth knowing: it usually makes your efficiency look better than it was, which is the opposite of the anxiety most people bring to their morning number. Devices also vary in how they define time in bed, so the same night can produce different percentages on different wrists. Read the two-week trend rather than the nightly figure, compare a device only against itself, and let your own experience of the night outrank the graph. Are sleep trackers accurate unpacks the validation research in more detail.
How do you improve sleep efficiency?
The most evidence-backed answer is counterintuitive: spend less time in bed.
Sleep restriction therapy — the component of cognitive behavioral therapy for insomnia that compresses time in bed to roughly match actual sleep, then expands it gradually as sleep consolidates — produces moderate-to-large improvements in sleep onset latency, wake after sleep onset, and sleep efficiency in reviewed trials, with only a small improvement in total sleep time (Miller et al., 2014, Sleep Medicine Reviews). In other words, it works mainly by making your sleep denser, not longer.
Two honest caveats. It is uncomfortable in the first weeks: restricting time in bed temporarily reduces objective total sleep time and increases daytime sleepiness before things improve, which is why it is best done with a clinician or a structured program rather than improvised. And it is not the right tool for fragmentation you do not control — a newborn, an untreated hot flash pattern, or pain will not respond to a narrower sleep window. How to treat insomnia without medication walks through the full CBT-I approach and who it suits.
Alongside that, three lower-stakes moves genuinely help the number:
- Move non-sleep activities out of bed. Reading, phone, television. This shrinks the denominator without costing you a minute of sleep.
- Hold your wake time steady. A fixed wake time anchors the body clock and is a stronger lever than a fixed bedtime — see sleep consistency vs duration.
- Stop going to bed early to “catch up.” Extra time in bed when you are not sleepy is the most common self-inflicted cause of a low number.
When should you ignore sleep efficiency entirely?
When the thing dragging it down is something you would not change even if you could.
A two-week-old who needs feeding. A period of illness. Grief. A partner on a different schedule. In those stretches, the number is simply reporting your life back to you, and turning it into a target adds a second problem on top of the first. There is good reason to be careful here: the anxiety of watching a sleep metric can itself fragment sleep, which is why we don’t show nightly scores in Mendtide.
And one clinical note. Persistently low sleep efficiency alongside loud snoring, witnessed breathing pauses, or heavy daytime sleepiness is worth raising with a doctor rather than solving with a routine change — those are the signs of sleep apnea, and no bedtime habit fixes them.
The calm version
Sleep efficiency measures how well your time in bed and your sleep line up — around 85 percent is the usual benchmark, and a low number most often means extra waking time in bed rather than anything wrong with you. In the newborn months and through perimenopause, a lower figure is the expected reading, and the trend across two weeks tells you far more than last night did. Mendtide treats numbers like this as context rather than a grade: patterns explained in plain language, with every claim traceable to the research behind it.
The percentage describes the fit between your bed and your sleep. It was never meant to describe you.