"Sleep divorce" is an ugly name for something quite ordinary: couples who decide to sleep in different beds, or different rooms, because that is the arrangement in which both of them sleep well.
The name makes it sound like a relationship failure with a mattress in the middle. The data is less dramatic than that. According to a survey by the American Academy of Sleep Medicine, nearly one-third of U.S. adults — 31% — have opted to sleep in another bed or another room at least some of the time, with adults aged 35 to 44 the most likely to do it (39%) and adults 65 and over the least likely (18%). That is not a fringe behaviour. It is a mainstream one that most people feel slightly awkward about.
So let us look at what the actual sleep research says, because the answer is genuinely two-sided, and the deciding vote is not cast by the internet. It is cast by your own nights.
TL;DR: Two large observational datasets and one sleep-lab study suggest that, on average, sharing a bed with a partner is associated with better sleep: about 10% more REM sleep and less REM fragmentation in a lab study of 12 couples, and less insomnia severity, less fatigue, shorter sleep onset and less time awake after falling asleep among the 1,007 adults in the SHADES study. But averages hide the couples who are being kept awake. If your partner snores heavily, moves constantly, or works a different schedule, the damage is measurable: bed partners of people with obstructive sleep apnea are around three times more likely to report insomnia symptoms, and in one early in-bed PSG study the wives of snoring men had a median sleep efficiency of just 74%. The honest answer is that sleeping together is better for most people most of the time, sleeping apart is a legitimate fix for specific problems, and the way to find out which you are is to run a proper two-week test with the data you already collect.
YouTube: Related video
What "sleep divorce" actually means
Two very different things get filed under the same term, and conflating them causes most of the arguments.
Sleeping in separate beds, same room. Twin beds, two mattresses pushed apart, one partner on the sofa for a bad night. You still share the room, the bedtime ritual and the last ten minutes of conversation. You remove the physical disturbance — the elbow, the duvet tug-of-war, the rolling mattress — but keep the noise and light.
Sleeping in separate rooms. Full separation. Two bedrooms, two thermostats, two alarm clocks, two different light environments. This is the version that solves snoring, opposite chronotypes, shift work and infant night feeds, and the version that partners worry means something.
There is also the timing question. Plenty of couples sleep together most nights and apart on specific ones — the night before an important day, when one of them is unwell, when one has been away travelling. That hybrid arrangement is probably the most common version of "sleep divorce" in practice, and the one least likely to bother anyone.
What the evidence says about sharing a bed
Start with the study that is quoted by both sides.
Researchers at Aarhus University and the University of Kiel put 12 young, healthy, heterosexual couples through four nights of polysomnography in a sleep laboratory — some nights together, some nights alone. The headlines that followed focused on one number: co-sleeping couples got about 10% more REM sleep, and their REM sleep was significantly less fragmented (p = 0.0006). Their limb movements increased during co-sleep, but their overall sleep architecture, sleep efficiency, awakenings and subjective sleep quality did not change. Couples also synchronised their sleep stages when sharing a bed, and the strength of that synchronisation tracked with how much significance they gave their relationship.
That is a genuinely pro-co-sleeping result, with caveats the authors were open about: 12 couples, all young and healthy, one lab setting, four nights. Small and specific.
The second dataset is bigger and points the same way, in the wild. The SHADES study surveyed 1,007 working-age adults in southeastern Pennsylvania about how often they shared a bed with a partner, a child, pets, other family, or no one. Compared with people who never shared a bed with a partner, those who did so most nights reported less insomnia severity, more total sleep, shorter sleep onset latency, less fatigue, less time awake after sleep onset and lower sleep apnea risk scores. They also scored lower on depression, anxiety and stress and higher on social support, life satisfaction and relationship satisfaction. Sleeping alone, by contrast, was associated with greater insomnia severity, more daytime sleepiness, more fatigue and higher depression scores.
And a smaller prospective study followed 29 co-sleeping couples with actigraphy and daily diaries for a week and found the relationship runs in both directions: men's sleep efficiency predicted their next-day negative interactions with their partner, while for women how closely their sleep onset matched their partner's predicted how they rated the following day, with greater discordance in sleep timing predicting more negative interactions.
Put together, the picture is consistent: on average, and in the absence of a specific sleep problem, sharing a bed is associated with better sleep and better mood — and sleeping alone is associated with worse sleep and lower mood.
That is the strongest card in the co-sleeping hand, and it is a real one. It is also an average, which is exactly where the argument gets interesting.
The case for sleeping apart
Averages are the wrong lens if you happen to be one of the people being woken up.
The clearest evidence comes from the sleep-apnea literature, where researchers have spent decades measuring not just patients but the people sleeping next to them. A review published in the Journal of Clinical Sleep Medicine pulled the numbers together:
- Compared with spouses from the general population, spouses of people with heavy snoring and obstructive sleep apnea are around three times more likely to report insomnia symptoms — difficulty falling asleep, staying asleep, and waking unrefreshed — and about twice as likely to report fatigue and daytime sleepiness, even after adjusting for age, BMI, number of children, work hours, their own snoring and sleep medication use.
- In one of the earliest in-bed studies, 10 wives of men undergoing sleep testing were monitored at the same time, in the same bed. Their median sleep efficiency was 74% with an arousal index of 21, and up to 32% of them had an arousal within one to three seconds of a snore from their husband. Their sleep was being converted into a series of micro-awakenings triggered by someone else's airway.
- Among 46 partners of patients with apnea, 69% reported moderate to severe sleep disturbance from snoring, 54% from observed apneas, and 55% from restlessness. Two-thirds were characterised as poor sleepers.
- In a group of 37 partners of men with suspected apnea, 35% reported relationship problems caused by snoring, and another study found wives of men with untreated apnea were 2.9 times more likely to sleep apart than wives of men who did not snore frequently.
Then there is the more subtle, universal problem: your partner wakes you and you wake less well than you think. A 2020 study in Sleep Health followed 52 couples aged 19 to 82 in which one partner had insomnia, and quantified "wake transmission". Averaged across couples, they described sleep concordance of about 67%, wake concordance of around 7%, and roughly six wake transmissions received per night — about 19% of a partner's awakenings crossed over, and couples spent around half their partner's wake minutes resistant to being woken. Couples with similar bedtimes (within 30 minutes) had the highest rate of transmissions received and the lowest resistance, which tells you something both useful and slightly funny: when you go to bed at the same time as your partner, you are maximally exposed to their night.
Finally, the mismatches that no amount of love fixes: opposite chronotypes, shift work and night feeds. A night owl and a lark sharing a bed are not just waking at different times. They are each other's light and noise source at the two most sensitive points of the 24-hour cycle, which is precisely where the sleep timing regularity evidence says the damage is done.
What actually changes when you sleep apart
If you are going to experiment, know what you are measuring. This is the honest accounting.
| Sleeping together | Sleeping apart | |
|---|---|---|
| Physical disturbance | Elbow, duvet, movement, mattress transfer | Removed |
| Snoring / apnea noise | Fully exposed | Removed |
| Light and alarm timing | Compromised unless aligned | Fully independent |
| Room temperature | Compromise | Each person optimises |
| REM sleep on average | About 10% more, less fragmented in lab data | Likely lower, if the lab pattern holds |
| Sleep-stage synchronisation | Present, tracks relationship depth | Absent |
| Falling asleep | Often faster | Reportedly can be slower for some |
| Emotional / relational | Intimacy, security, shared wind-down | Requires deliberate replacement |
Note the asymmetry. The costs of sleeping apart are mostly social and psychological; the costs of sleeping together when there is a real disturbance are physiological and measurable. That is why the right answer depends less on principle and more on whether the disturbance in your bed is above or below the threshold where it costs you sleep.
How to test it on yourself in two weeks
Here is the part no headline can do for you. You have a wearable that has been quietly recording both arrangements for years.
A clean test looks like this:
1. Two weeks, alternating, and not one night each. Sleep in separate arrangements for at least three consecutive nights in each condition, ideally four or five. The reason is the first-night effect — the first night in any new bed or new arrangement is measurably worse, which will make a one-night test nonsense. Give each arrangement long enough to get past novelty.
2. Do not change anything else. Same bedtime target, same caffeine cutoff, same alcohol, same training load, same screen habits. You are isolating one variable. If you also start going to bed an hour earlier, you have learned nothing.
3. Read four numbers, and read them as trends:
- Time awake after sleep onset and awakening count. This is the number that moves first when someone is disturbing you. If it drops clearly and consistently when you sleep apart, you have your answer.
- Sleep efficiency. The percentage of time in bed actually spent asleep. A jump from 82% to 90% across two weeks is a real, felt difference.
- Sleeping heart rate. If your overnight heart rate is consistently lower on the nights you sleep alone, your body is telling you the shared bed was costing you autonomic recovery.
- HRV. Use the trend, not individual nights. A clear multi-night shift in overnight HRV alongside the other three is the strongest signal you will get from consumer hardware.
- Bonus: sleep consistency. Sleeping apart often fixes bedtime irregularity as a side effect, which shows up as a tighter window of sleep onset times across the fortnight.
4. Record how you feel, briefly. One or two words the morning after each night. Sleep data does not capture the part where you missed them.
Reading the result honestly
Most couples who run this test land in one of three places.
"The data barely changed." Very common. If sleep efficiency, awakenings, sleeping HR and HRV are stable across both arrangements, then your partner is not disturbing you, and the co-sleeping benefits in the research — the emotional regulation, the faster sleep onset, the synchronisation — are the ones you are actually getting. Sleep together.
"Everything improved, and I want to keep doing it." Also common, and it usually has a physical cause: snoring, movement, opposite schedules or a small bed. Sleep apart, at least on the nights that matter. This is not a verdict on the relationship. It is a fix for a mechanical problem, and the person snoring should still get assessed — our guide to what snoring actually means covers the red flags worth raising with a doctor, and treating the airway is often what saves the shared bed.
"I slept better alone but I hated it." This is the most common outcome of all, and the most informative. The answer for these couples is rarely all-or-nothing. It is usually part-time.
Middle grounds worth trying before you move bedrooms
- Two duvets, one bed. The single highest-yield, lowest-cost intervention in shared sleeping. Most disturbance is temperature and covers, not bodies.
- A bigger bed. Obvious, boring, effective. Going from a double to a king removes a surprising amount of wake transmission.
- White noise on the snorer's side. Masks the peaks of snoring and reduces the fine-grained micro-arousals that never fully wake you but still fragment sleep.
- A part-time schedule. Together on weekends and free mornings, apart on the night before anything that matters. Many couples converge on this and never call it anything.
- Treat the actual problem. CPAP, an oral appliance, positional therapy, nasal surgery, treating restless legs, treating insomnia properly. The bed-partner literature is clear that treating the apnea improves the partner's sleep too — which is the outcome you actually want.
- Fix bedtime mismatch where you can. You will not turn a night owl into a lark. But a shared wind-down in a dim room costs little and buys the synchronisation benefit without forcing the same sleep onset.
If you go for separate rooms, treat the intimacy as a schedule problem rather than something you have lost. The couples who handle this best are usually the ones who replaced the falling-asleep-together ritual with something deliberate earlier in the evening.
When it is not about sleep at all
One honest caveat. Sleeping apart is a sleep decision that can also become a relationship decision, and the research here is more suggestive than conclusive: sleep and relationship functioning run in both directions, with poor sleep predicting more negative interactions the next day and relationship conflict predicting worse sleep that night.
So use this as a test: if sleeping apart is making both of you better-rested, patient, and easier to be around, it is a sleep intervention that is helping the relationship. If it is a way of avoiding someone, the improved sleep numbers will not save you from the rest of it.
And be sceptical of a single bad night. In a shared bed, no one sleeps perfectly every night — that is true of sleeping alone too. That is the same principle behind why trends matter more than daily scores, and it is exactly what Century AI is designed for: a daily health score, recovery score and sleep insight built from the watch you already wear, so you can see a fortnight of nights as a pattern instead of a verdict. If you are going to make a decision this consequential about where you sleep, make it on two weeks of nights, not three.
The bottom line
Sleep divorce sounds like a crisis. It is usually just logistics. Across large surveys and lab studies, sharing a bed is associated with better sleep on average, more REM sleep and less insomnia — which is a real argument for staying put. But when one partner has untreated snoring or apnea, a very different schedule, or genuinely disruptive movement, the measured cost to the other person is large enough to justify sleeping apart, and often large enough to make it the healthier choice for the relationship.
You do not have to guess. Run the fortnight, read your own awakenings, sleeping heart rate and HRV as trends, and let the data decide which side of the average you are on. Then keep the parts of the arrangement that were actually working — the wind-down, the conversation, the part where you like each other.
Century AI helps you understand your body with a daily health score, recovery score, and sleep insights — using the watch you already wear.