The First-Night Effect: Why You Sleep Badly in a New Place

The effect itself has been documented since the 1960s and is so reliable that sleep laboratories routinely discard the first night's data as unusable. The interesting part is the explanation, which arrived only in 2016.

What the effect looks like

In an unfamiliar sleeping environment, polysomnography shows a consistent pattern: reduced total sleep time, more frequent awakenings, longer time to reach REM sleep, and less REM overall. Sleep efficiency drops. By the second night in the same place, most of this has resolved.

This is why sleep studies build in an adaptation night. Researchers spent decades treating the first-night effect as a nuisance to be discarded before anyone asked what it was for.

The night watch study

Masako Tamaki, Yuka Sasaki and colleagues at Brown University ran three experiments across 35 volunteers, each sleeping in the laboratory on two nights a week apart, using magnetoencephalography combined with structural imaging to measure activity network by network and hemisphere by hemisphere.

On the first night, slow-wave activity — the standard index of sleep depth during non-REM sleep — was lower in the left hemisphere than the right, specifically within the default mode network. The size of that asymmetry correlated with how badly the person slept: more asymmetry, stronger first-night effect. On the second night the asymmetry was gone.

The team then tested whether the shallower hemisphere was actually doing anything. Sleeping participants were played irregular deviant tones among regular beeps. On the first night, the left hemisphere produced larger evoked responses to the deviant sounds than the right. In a third experiment, participants were asked to tap a finger on hearing a sound; deviant tones delivered to the right ear — and therefore primarily to the left hemisphere — produced faster arousal and more frequent responses than the same tones delivered to the left ear.

So the shallower hemisphere is not merely less asleep. It is more responsive to exactly the kind of stimulus that would matter if something were wrong.

What this is not

Headlines described half the brain staying awake. That is not what was found, and the gap matters.

The asymmetry was regional, confined to one network rather than a whole hemisphere. It was a difference in the depth of slow-wave sleep, not a difference between sleep and wakefulness. And it was invisible to standard methods — the researchers noted that visual inspection of the polysomnography showed no hemispheric asymmetry, and that sensor-space frequency analysis failed to reveal it either. It took network-level analysis across two nights to see at all.

True unihemispheric sleep, in which one hemisphere shows full waking EEG while the other sleeps, occurs in dolphins, some other cetaceans, and several bird species. Humans do not do this. What humans appear to do is a much subtler version: a modest, regionally specific reduction in sleep depth on one side, coupled with heightened responsiveness.

The functional logic is familiar, though. A sleeping animal is defenceless, so a system that lowers its detection threshold in unfamiliar surroundings buys safety at the cost of sleep quality — the same asymmetric-cost reasoning that keeps your face-detection threshold low enough to find faces in wall sockets. False alarms are cheap. Missing the real thing is not.

How firm is it?

The first-night effect is beyond dispute — six decades of laboratory data. The night-watch explanation is a different matter and deserves proportionate confidence.

Thirty-five participants across three experiments means roughly a dozen each. The finding is internally consistent and the convergence across three different measures is impressive, but it comes from one group and has not been independently replicated at scale. The same team published follow-up work extending the surveillance account to REM sleep in 2019, which strengthens the picture without adding independent confirmation.

The left-hemisphere specificity is also unexplained. Nothing in the account predicts which side should take the watch, and whether that would hold across a larger and more varied sample is unknown.

Who does not get it

The effect is not universal, and the exceptions are informative. A reverse first-night effect has been documented in some people with insomnia, who sleep better in a sleep laboratory than at home — plausibly because the bedroom itself has become the conditioned trigger for arousal, so a neutral unfamiliar room is an improvement.

That inversion fits the surveillance account rather than contradicting it. What drives the response is the brain's assessment of the environment, and for someone whose own bedroom has accumulated years of association with lying awake, the unfamiliar room is the safer one.

Why it fades

By the second night in the same room, the asymmetry disappears and sleep normalises. The environment has been sampled, nothing bad happened, and the threshold returns to baseline — an unusually clean demonstration of a threat-assessment system updating on evidence.

This also explains the everyday pattern. The effect is strongest in genuinely unfamiliar places, weaker in a hotel chain you have stayed in before, and absent at home. What matters is novelty rather than comfort: an expensive unfamiliar bed still triggers it and a mediocre familiar one does not.

It fits the broader picture of sleep as an actively managed state rather than an off switch — the same capacity that produces a hormonal ramp timed to an alarm you set before falling asleep. The sleeping brain is doing considerably more monitoring than the experience of sleeping suggests.

Anything to do about it?

Nothing dramatic, and the honest answer is that the mechanism does not suggest an intervention. Reducing novelty is the only lever the account implies — familiar bedding, a familiar routine, sleeping in the same room on a repeat visit — and none of this has been tested as a way of suppressing the effect.

The more useful reframe is that a bad first night in a hotel is a normal, temporary, self-correcting response to an unfamiliar environment rather than a sign that something is wrong with your sleep. It generally sorts itself out by the second night without anyone doing anything.