There is no single correct bedtime, but there is a reasonably good way to find yours: work backwards from the time you actually have to wake up. Subtract the amount of sleep your body needs — for most adults that’s 7 to 9 hours — then subtract another 15 to 20 minutes for the time it takes to fall asleep. If you have to be up at 6:30am and you need eight hours, your bedtime is somewhere around 10:10pm. That’s the arithmetic. What the research adds is that the consistency of that bedtime tends to matter more than the exact number on the clock.
For two groups in particular — new parents, whose nights are broken by someone else’s schedule, and women in perimenopause, whose internal clock is quietly shifting underneath them — the standard bedtime advice tends to land badly. So it’s worth separating what the evidence actually says from what gets repeated.
How do I calculate my own bedtime?
Three inputs, in this order:
1. Your fixed wake time. Not your ideal one — the real one, the one an alarm, a commute, or a child enforces. This is the anchor, because wake time plus morning light is what holds your circadian clock in place.
2. Your sleep need. A consensus statement from the American Academy of Sleep Medicine and the Sleep Research Society recommended 7 or more hours per night for adults aged 18 to 60 (Watson et al., Sleep, 2015), while the National Sleep Foundation’s expert panel landed on a 7-to-9-hour range for the same group (Hirshkowitz et al., Sleep Health, 2015). Individual need varies within that band; our guide to how much sleep you need covers how to estimate yours.
3. Sleep latency. Healthy sleepers typically take somewhere in the range of 10 to 20 minutes to fall asleep. Build that into the calculation rather than expecting to be unconscious the moment your head lands.
So: wake time, minus sleep need, minus twenty minutes. That’s your target lights-out. Treat it as a window of about half an hour rather than a precise minute — precision here creates anxiety without adding benefit.
Is 10pm actually the best bedtime?
This is the claim you’ve probably seen, and it comes from a real study worth understanding properly.
Nikbakhtian and colleagues analysed accelerometer data from more than 88,000 UK Biobank participants, using wrist-worn devices to capture actual sleep onset rather than self-reported bedtime. Over roughly six years of follow-up, sleep onset between 10pm and 11pm was associated with the lowest incidence of cardiovascular disease. Falling asleep at midnight or later carried a higher associated risk, and — notably — so did falling asleep before 10pm (Nikbakhtian et al., European Heart Journal – Digital Health, 2021).
Two caveats keep this honest. First, it’s observational: people who fall asleep at 1am differ from people who fall asleep at 10:30pm in many ways beyond bedtime, and the study can’t fully untangle those. Second, the association was clearer in women than men, and the mechanism proposed — circadian misalignment — implies that what matters is the gap between your body clock and your schedule, not the number 10 itself. A genuine night owl forced into a 10pm bedtime is arguably creating misalignment, not resolving it.
The useful takeaway isn’t “go to bed at 10.” It’s that habitually falling asleep well after midnight, on a schedule your body didn’t choose, is the pattern associated with worse outcomes.
Why does bedtime consistency matter more than the exact hour?
Because irregularity appears to carry its own cost, independent of duration.
Windred and colleagues examined roughly 60,000 UK Biobank participants with wearable data and found that sleep regularity — how similar your sleep-wake timing is from one 24-hour period to the next — was a stronger predictor of mortality risk than sleep duration (Windred et al., Sleep, 2024). Earlier work pointed the same direction: Phillips and colleagues, publishing in Scientific Reports in 2017, found that students with irregular sleep patterns had delayed melatonin release and poorer academic performance than regular sleepers even when total sleep time was the same (Phillips et al., Scientific Reports, 2017). And Lunsford-Avery and colleagues found sleep irregularity associated with higher cardiometabolic risk in older adults (Lunsford-Avery et al., Scientific Reports, 2018).
Roenneberg’s team named the everyday version of this social jetlag: the gap between your sleep timing on work days and free days, which behaves a little like flying across time zones every weekend (Wittmann et al., Chronobiology International, 2006). A two-hour weekend drift is common and worth narrowing. Zero drift is not a realistic target for most people.
If you want the longer version of this trade-off, we’ve written about sleep consistency versus duration in more detail.
What if I’m a night owl?
Then your calculated bedtime may sit an hour or two later than the general advice, and that’s a fact about your biology rather than a discipline problem.
Chronotype — your internal preference for earlier or later timing — is substantially heritable and forms a broad bell curve rather than two camps (Roenneberg et al., Current Biology, 2004). It also shifts with age, running latest around age 20 and drifting earlier through adulthood. Our guide to chronotype and sleep walks through how to identify yours using your natural sleep midpoint on days with no alarm.
The practical move for a late chronotype isn’t to force an earlier bedtime by willpower. It’s to shift the whole clock gradually, using morning light and a fixed wake time as the levers, in increments of 15 to 30 minutes. How to fix your sleep schedule covers that process step by step.
What bedtime makes sense for new parents?
Honestly? For the first several months, “bedtime” as a single fixed point may not exist — and pretending otherwise just adds a target to fail at.
Montgomery-Downs and colleagues tracked maternal sleep across the first four postpartum months and found total sleep time was not dramatically reduced, but sleep was heavily fragmented, with substantially more time awake after sleep onset (Montgomery-Downs et al., American Journal of Obstetrics and Gynecology, 2010). Fragmentation, not duration, is the defining feature.
Three adjustments that tend to hold up:
- Anchor the earliest possible sleep opportunity. If the reliable stretch is 9pm to midnight, protect it. Going to bed when the baby does, even at an hour that feels absurdly early, buys the deepest sleep of the night — slow-wave sleep is front-loaded.
- Let wake time float more than you normally would. During fragmented months, rigid wake times cost more than they return.
- Stop treating “catching up” as failure. Recovery sleep is real, if incomplete — see sleep debt for parents for what it can and can’t repay.
What about bedtime in perimenopause?
The complication here is the opposite one: the pull is toward going to bed too early.
The circadian system phase-advances with age, so sleepiness arrives earlier in the evening while the clock still releases you at the same early hour — which is why 9pm sleepiness and 4:30am waking often travel together. Sleep in the menopausal transition is also disrupted by vasomotor symptoms, with hot flashes and night sweats fragmenting the first half of the night (Baker et al., Sleep Medicine Clinics, 2018).
The trap is compensating for a poor night by getting into bed at 8:30pm. That extends time in bed without extending sleep, and typically produces an even earlier final waking. If early waking is the main symptom, holding a slightly later bedtime is usually the more effective adjustment — waking up too early and perimenopause and sleep both go further into why.
What if I’m not sleepy at my calculated bedtime?
Then don’t go to bed yet. This is the single most common way a “correct” bedtime backfires.
Borbély’s two-process model describes sleep as governed by two systems working together: homeostatic sleep pressure, which builds steadily the longer you’re awake, and the circadian process, which gates when that pressure can be discharged (Borbély, Human Neurobiology, 1982). Getting into bed early doesn’t add sleep pressure — it just gives you more time to lie there noticing you’re awake, which is precisely the association that makes insomnia self-sustaining.
Clinically, the counterintuitive fix is to compress time in bed rather than expand it: sleep restriction therapy, described by Spielman and colleagues in Sleep in 1987, deliberately shortens time in bed to consolidate sleep, then extends it as efficiency improves. You don’t need the full protocol to use the principle. If you’re regularly spending 30-plus minutes awake at lights-out, your bedtime is probably too early, not too late. Tired but can’t sleep unpacks that mismatch further.
The calm version
Your bedtime is a subtraction problem, not a moral one: fixed wake time, minus your sleep need, minus about twenty minutes. Keep it within a rough half-hour window most nights and you’ve captured nearly all of the available benefit — the regularity matters more than the specific hour, and the specific hour matters less than whether it fits your own chronotype. If you’re deep in newborn months or navigating perimenopausal early waking, the honest answer is that the standard formula bends, and that’s expected rather than a failure. Mendtide reads your actual sleep timing from Apple Health and shows you where your real bedtime sits versus the one you intended, without grading you on the gap.
The right bedtime isn’t the one that sounds virtuous. It’s the one your body can actually fall asleep at, roughly the same time, most nights.