Why Sleepwalking Happens in the First Half of the Night

01The clock you noticed is a real clock

Start with what you saw, because you were reading it correctly. Sleepwalking has a time of night, and it is early.

MedlinePlus states it plainly: sleepwalking "most often occurs during deep, non-REM sleep (called N3 sleep) early in the night." Mayo says the same thing with a number attached, one to two hours after falling asleep, and adds that a bout "generally lasts several minutes, but it can last longer." MedlinePlus puts most episodes under ten minutes.

Those two facts, an early window and a short duration, are the shape of almost every account a parent or a partner gives. Somebody appears, does something semi-purposeful for a few minutes, and is back in bed before anyone has decided what to do. The next morning there is nothing to discuss, because the person it happened to has no idea it happened.

What almost nobody tells you is why the window sits where it does, and the gap is a strange one, because the causes are the uncertain part and the timing is the part we understand well. Those two get run together constantly, so separate them before going further.

Why some people sleepwalk and others never do is genuinely unsettled. There is a strong family pattern, there are known conditions that make episodes more likely, and there is no single mechanism anyone can point to and call the cause. Clinical sources say so openly, and they are right to.

Why episodes happen early in the night is not unsettled at all. It follows from how a night of sleep is built.

02A night of sleep spends its deep sleep first

Sleep is not one flat state that lasts eight hours. It runs in cycles, and the cycles are not identical to each other.

The National Heart, Lung, and Blood Institute describes the cycle as starting over every 80 to 100 minutes, with four to six cycles in a typical night. Each cycle moves through lighter non-REM sleep, into the deepest non-REM stage, and then into REM sleep, the stage where most vivid dreaming happens.

The important part is what changes from one cycle to the next. The NHLBI says of deep sleep, also called slow wave sleep, that "you usually spend more time in this stage early in the night," and of the other end, that "you usually have more REM sleep later in the night." A standard physiology summary puts the same pattern in one sentence: the first REM period is short, and as the night goes on there are longer periods of REM and less time in deep sleep.

So, a night has a direction to it. It begins heavy and ends light. The deepest sleep you will get is mostly spent in the first couple of hours, and by the last stretch before your alarm the mix has flipped almost entirely the other way, into long REM periods and dreams you might actually remember.

Two honest caveats about that table, because sleep architecture is easy to overclaim. The boundaries are not crisp lines on a chart; deep sleep tapers rather than stopping at a particular minute. And individual variation is large. Age matters a great deal, since deep sleep declines substantially across a lifetime, and so do how short of sleep you were beforehand, what time you went to bed, and whatever else is going on in your nights. A table like this is the shape of a typical night, not a schedule your body has agreed to.

But the direction of travel is reliable, and it is enough to answer the question. Anything that arises specifically out of deep sleep has most of its opportunities early, because that is when nearly all of the deep sleep is.

03What is actually happening in those few minutes

Here is the piece that makes the rest of it make sense, and it is more interesting than the tidy version.

A sleepwalking episode is not a person waking up and doing something odd. It is not a person unconscious and moving on autopilot either. It is a state that has genuinely no waking equivalent, and clinical literature has a precise phrase for it. Sleepwalking, sleep terrors and confusional arousals are described as resulting from an incomplete dissociation of wakefulness from non-REM sleep, in an updated review by Irfan, Schenck and Howell in Neurotherapeutics. Wakefulness and deep sleep are supposed to be mutually exclusive. In an episode they are not, and both are running at once in different parts of the same brain.

Which parts do which turns out to be the whole explanation. A 2022 review by Idir, Oudiette and Arnulf in the Journal of Sleep Research gathers the behavioral, cognitive and brain recording evidence and describes a dissociated pattern during episodes that fits the newer idea of local arousal: wake-like activation in motor and limbic regions, alongside preserved, or even increased, sleep intensity over a frontoparietal network.

Read that slowly, because it answers three separate questions at once.

The regions that move the body are behaving as though awake. That is why somebody can stand, walk, open a door, navigate stairs and carry out something that looks purposeful. The motor system does not need you to be conscious to run; it needs to be switched on.

The emotional regions are also behaving as though awake. That is why episodes so often carry an atmosphere of urgency or fear rather than a calm stroll, and why a child can be crying while nothing has actually happened to them.

And the frontoparietal network, which is a large part of how a person is aware of their situation, weighs a decision, recognizes a face and lays down a memory of any of it, is not merely asleep. It is sleeping harder than usual. That is why the person in front of you can be looking straight at you and not be reachable, and why nothing about the episode survives until morning.

Now put that back on the clock, because it is the same fact from the other side. If an episode requires being deep in slow wave sleep and coming partly out of it, then it requires slow wave sleep to be there in the first place. Early in the night there is a great deal of it. Late in the night there is barely any. The window is not a mystery once you know what has to be present for the door to open.

There is even direct evidence that the mistiming is the point. Cataldi and colleagues recorded 20 adults with non-REM parasomnias using high density EEG, then compared 96 parasomnia episodes against 25 ordinary awakenings in the same people. Both were preceded by the same thing: large, steep slow waves in frontal and central regions together with a rise in fast brain activity. That process is how anybody surfaces out of deep sleep. The difference was that in the parasomnia episodes it arrived on a less activated background, during a period of high slow wave activity, and slow wave activity stayed higher after the movement began. Their conclusion is that an abnormal timing of that arousal process could underlie these events.

That is a genuinely reassuring finding, and it is worth saying out loud. The thing happening in your hallway is not an alien process. It is the same waking-up machinery everyone has, firing at the wrong moment in the wrong depth of sleep, and then only getting halfway.

04Why there is no memory of it, and why the eyes are open

The blank next morning unsettles people more than the walking does, particularly when the sleeper is an adult being told what they did. It helps to know that the blank is the expected result rather than a symptom of anything extra.

Memories of an event are not recorded automatically by being present. They are built, by systems that were, in this case, in deep sleep the entire time. Nothing was written down, so there is nothing to retrieve. Mayo lists not remembering in the morning among the ordinary features, alongside sitting up with the eyes open, a glazed expression, not responding when spoken to, being hard to wake, and being confused for a short time afterward.

The open eyes catch people off guard every time. Someone whose eyes are open, who is upright and moving, reads to us as awake, and every instinct says to talk to them. They are not awake, and the part of them that would process what you said is the part that is furthest under. That is not a failure to try hard enough on your end.

This is also where the one comparison worth making belongs, said once and then left alone. Behavior that comes out of REM sleep, where somebody physically acts out a dream they can describe afterward, sits at the opposite end of the night for exactly the reason this whole page has been describing: REM is concentrated in the final third, which is where deep sleep is not. If that is closer to what you are seeing, our page on acting out dreams while asleep is the one that covers it properly, and it belongs to a different conversation than this one.

Which leaves an obvious question. If the sleeper is not enacting a dream, was there anything in their head at all?

The honest answer is more interesting than a flat no. Oudiette and colleagues, writing in Sleep in 2009, interviewed 43 adults with sleepwalking or sleep terrors alongside 25 healthy controls and found that "short, unpleasant dreamlike mentations may occur during sleepwalking/sleep terrors episodes, suggesting that a complex mental activity takes place during slow wave sleep." They go as far as saying sleepwalking may represent acting out of that mentation. And a 2021 study by Castelnovo and colleagues interviewed 45 patients, 25 adults and 20 children, and found a striking split by age: more than two thirds of the children could recall no mental activity at all, while more than two thirds of the adults recalled at least one mental experience, often brief and often carrying apprehension, set in their own home.

So, not a REM dream being performed, and not a blank either. Usually a fragment, closer to a single alarming image than to a story, and more likely to be reported by an adult than by a child. If you are a parent whose seven year old remembers nothing whatsoever, that is exactly what the research would predict.

05Why tonight and not last night

Once the timing makes sense, the next question is always the trigger. If deep sleep happens every night, why does an episode happen on some nights and not others?

The most useful answer available comes from asking sleepwalkers themselves in large numbers. Blanchette-Carrière and colleagues surveyed 188 adults diagnosed with primary sleepwalking, published in the Journal of Sleep Research, about what they perceived as contributing to their episodes at home rather than in a laboratory.

Two things there deserve emphasis. Stress is the standout by a wide margin, and it is the item most likely to be missing from a list of medical causes. And substances, where popular explanation goes first, sit near the bottom rather than the top. These are self-reports, so they describe what people believe about their own nights rather than proving cause, but 188 diagnosed sleepwalkers noticing the same pattern is worth more than an anecdote.

Sleep loss also has something firmer behind it than a survey. Zadra, Pilon and Montplaisir evaluated 30 consecutive sleepwalkers by video sleep study, on a normal baseline night and again during recovery sleep after 25 hours awake, publishing in Annals of Neurology in 2008. Pooling all 40 patients once people with an additional sleep disturbance were included, 32 episodes were recorded from 20 sleepwalkers at baseline, and recovery sleep produced 92 episodes from 36 patients. Their interpretation is that sleepwalkers have trouble sustaining stable slow wave sleep and are particularly vulnerable to increased sleep pressure.

That is the mechanism from earlier, described in a different vocabulary. Miss a night of sleep and your body answers by pushing deep sleep harder on the next one. More deep sleep, and more pressure inside it, means more chances for an incomplete arousal to occur.

The other half of it is interruption, and there is a companion study on that too. Pilon, Montplaisir and Zadra tested 10 sleepwalkers and 10 controls with sounds delivered during slow wave sleep, on normal sleep and on recovery sleep. No episode was ever induced in a control. Among the sleepwalkers, all of them had at least one induced episode during recovery deep sleep, against 30 percent during normal deep sleep.

The detail worth carrying home from that study is a quiet one. The sounds that triggered episodes were no louder, on average, than the sounds that simply woke people up. It was not a matter of an unusually big disturbance. An ordinary interruption, arriving at the wrong depth in the wrong person, was enough.

Which is why a hallway light, a door, a full bladder, a snoring partner, a fever, or a new and unfamiliar place can all sit in the background of a night that produced an episode. None of these is a cause you failed to eliminate. They are ordinary features of ordinary nights, and they matter more in the early hours because that is when there is deep sleep for them to interrupt.

Medicines deserve a careful mention rather than an alarming one. Stallman, Kohler and White reviewed 62 papers on medication and sleepwalking, identifying 29 drugs across four broad classes as possible triggers, and found the strongest evidence for zolpidem and sodium oxybate, with everything else resting on case reports. If episodes began around a new prescription or a dose change, that is genuinely useful information and it belongs with the prescriber, along with the dates. It is not a reason to stop or adjust anything on your own.

06Is this normal, and does it stop

For most households reading this, the honest answer is reassuring, and it is worth having the actual numbers rather than a vague sense that it is common.

Stallman and Kohler pooled 51 studies covering 100,490 people in a meta-analysis published in PLOS ONE. Lifetime prevalence of sleepwalking came out at 6.9 percent. Prevalence within the previous twelve months was significantly higher in children, at 5.0 percent, than in adults, at 1.5 percent. The authors are careful to note a significant risk of bias across the studies they pooled, so treat these as the best available estimates rather than exact figures.

The childhood picture is clearer still in a long follow-up. Petit and colleagues tracked 1,940 children born in Quebec from toddlerhood to age 13, publishing in JAMA Pediatrics. Sleepwalking prevalence peaked at age 10, at 13.4 percent, and sleep terrors peaked far earlier, at 18 months, at 34.4 percent. Around a third of children who had early sleep terrors went on to sleepwalk later in childhood, which is a useful thing to know if your household has been through both. Mayo's summary of the trajectory is the plain one: children usually outgrow sleepwalking by the teenage years.

The family pattern in that study is strong enough to be worth naming, because parents often suspect it before anyone tells them. Childhood sleepwalking ran at 22.5 percent among children with no parental history, 47.4 percent where one parent had sleepwalked, and 61.5 percent where both had. If you are reading this because you recognize your own childhood in your child's hallway, you are not imagining the connection.

For adults, the most useful finding is about continuity rather than onset. Hublin and colleagues studied 11,220 Finnish adults from a twin cohort, published in Neurology, and found that among adults who sleepwalked, 88.9 percent of men and 84.5 percent of women had a history of it in childhood. People who reported never sleepwalking as children rarely did as adults, at 0.6 percent.

That distinction is worth holding onto. If this has been part of your life since you were small, it is behaving the way the research says it behaves. If it is genuinely new in adulthood, that is a different conversation and a good reason to take it to a clinician rather than to a search bar.

07What a predictable window actually lets you do

This is where the timing stops being a curiosity and starts being useful, and it is the practical payoff of everything above.

You do not have to watch the whole night. If episodes cluster in the couple of hours after sleep onset, then that is the stretch that needs to be safe, and it is also the stretch during which somebody in the house is most likely to still be awake anyway. That is a much smaller problem than the one most people think they have been handed.

The safety work is unglamorous and effective. Doors to stairs and to outside matter most, since falls and leaving the house are where real injuries come from, and keys belong somewhere that takes a waking brain to find. Nothing sharp, hot or breakable in the path between the bed and the bathroom. Windows secured, particularly in a child's room. Clutter off the floor in the route they usually take, because they will take it again. None of this needs to be a fortress. It needs to remove the two or three things that could turn an ordinary four minute episode into an injury.

The in-the-moment part is simpler than people expect. Guiding gently back to bed, without questions and without trying to explain anything, is the default and it usually works, because the motor system will follow a light steering hand while the rest of them is not available for conversation. On the folklore about waking someone, MedlinePlus is unambiguous: "It is not dangerous to awaken a sleepwalker, although it is common for the person to be confused or disoriented for a short time when they wake up." So, if guiding is not working, or if the situation has become genuinely unsafe, waking them is not the thing you have been warned it is. Confusion afterward is the cost, and it is a small one.

One more option follows directly from the predictability, and it is rarely mentioned outside clinics. If episodes reliably arrive at around the same time, some families are taught to briefly rouse the sleeper roughly fifteen to twenty minutes beforehand, for a few nights running, so the deep sleep is interrupted before an episode can build out of it. A 2023 systematic review by Mundt and colleagues in Sleep Medicine gathered 72 publications on behavioral treatments for non-REM parasomnias and lists scheduled awakenings among the approaches with some supporting evidence.

Be honest about how strong that evidence is, though, because the same review is. Most of the 72 publications were case reports or case series, and the authors say directly that study designs and inconsistent outcome measures limit the evidence for any specific treatment. It is a technique to raise with a clinician who knows the person in question, not a protocol to start tonight because a website mentioned it.

The last thing the timing tells you is what to write down. The fields that earn their place are bedtime, the time the episode started, roughly how long it lasted, what they did, whether they left the room, how much sleep they had the night before, and anything unusual about the day. Two weeks of that is worth more at an appointment than any amount of retelling, because the person it happened to cannot contribute a single detail.

08When it is worth a clinician

Most sleepwalking never needs an appointment. Mayo's own line is that occasional bouts are not usually a cause for concern and can simply be mentioned at a routine check-up.

There is a shorter list of situations where a conversation is genuinely worth having, and it is a list of circumstances rather than symptoms to audit yourself against. Mayo suggests seeing a healthcare professional when episodes happen often, more than once or twice a week or several times in one night; when they lead to dangerous behavior or injury to the sleeper or anyone else; when they are badly disrupting the sleep of the household; when they leave someone very tired during the day or struggling at school or work; and when they continue into a child's teenage years or begin for the first time in adulthood.

Two additions are worth making to that. Loud snoring, gasping, or witnessed pauses in breathing alongside episodes is worth mentioning specifically, because sleep-disordered breathing is one of the conditions that can fragment deep sleep and is treatable in its own right. And if episodes started around a new medicine or a dose change, bring the dates, and let the prescriber be the one who decides what happens to the prescription.

None of that is a reason to sit up tonight worrying. It is a reason to make the appointment during business hours, with a couple of weeks of notes in hand.

09Common questions about sleepwalking and the time of night

Why does sleepwalking happen early in the night?

Because sleepwalking arises out of deep, slow wave non-REM sleep, and most of your deep sleep is concentrated in the first part of the night. The NHLBI describes cycles repeating every 80 to 100 minutes with more deep sleep early and more REM sleep later, so the raw material for an episode is abundant in the first couple of hours and largely gone by morning.

Can sleepwalking happen later in the night?

It can, and some households do see episodes closer to one or two in the morning, because deep sleep has not disappeared entirely by then. It becomes progressively less likely as the night goes on and the balance shifts toward REM sleep. Behavior in the last stretch before waking, particularly if the person can describe a matching dream afterward, is more likely to be a different kind of event.

Is a sleepwalker dreaming?

Not in the way people usually mean. This is not a REM dream being physically performed. But research interviewing sleepwalkers has found brief, often unpleasant dreamlike fragments reported around episodes, and adults report them far more often than children do. So, there is frequently something in mind, and it is usually a fragment rather than a story.

Why does the person not remember anything?

Because the parts of the brain that build memories were deeply asleep throughout. Recording evidence shows sleep intensity preserved or even increased over the frontoparietal network during episodes, at the same time as movement regions behave as though awake. Nothing was written down, so there is nothing to recall, and a complete blank is the expected outcome rather than a worrying extra.

Why do their eyes look open if they are asleep?

Open eyes and a glazed expression are among the ordinary features Mayo lists, along with not responding when spoken to and being hard to wake. It is one of the more unsettling things to witness, because everything about it reads as awake, but the systems that would let someone recognize you and answer are the ones most deeply asleep.

Does missing sleep make an episode more likely?

In people who already sleepwalk, yes, and this is one of the better supported findings in the field. In a video sleep study of 40 patients, 32 episodes were recorded from 20 people at baseline, while recovery sleep after 25 hours awake produced 92 episodes from 36 people. Sleep loss increases the pressure and depth of the following night's slow wave sleep, which is the state episodes emerge from.

Is it dangerous to wake a sleepwalker?

No. MedlinePlus states that it is not dangerous to awaken a sleepwalker, although confusion or disorientation for a short time afterward is common. Gently guiding someone back to bed is usually the easier route and avoids that confusion, but if guiding is not working or the situation has become unsafe, waking them is a reasonable thing to do.

Will a child grow out of sleepwalking?

Usually. Prevalence within the last year runs at about 5.0 percent in children against 1.5 percent in adults, childhood sleepwalking peaks around age 10, and Mayo's summary is that children generally outgrow it by the teenage years. Adults who sleepwalk overwhelmingly did so as children too, and people who never sleepwalked in childhood rarely start later.

Does stress cause sleepwalking?

Stress does not appear to create the underlying susceptibility, but it is by far the most commonly reported contributor among people who already have it. In a survey of 188 diagnosed adult sleepwalkers, 95 percent endorsed psychological stress as a precipitating or priming factor, ahead of bad dreams at 78 percent and sleep deprivation at 60 percent. That is self-reported rather than experimental, and it is still the clearest pattern in the data.

If you came here because the clock kept repeating itself, the answer is that your household has been watching ordinary sleep physiology run on schedule. Deep sleep comes first, an episode is a partial exit from it, and the window closes on its own as the night turns toward morning. Why any one person is susceptible is still an open question, and anybody who tells you otherwise is overselling. The timing is not open at all, and it is the part that lets you do something useful: make one specific hour of the night safe, keep a short record, and take that record to somebody if the situation calls for it. DreamTold treats dreams as a window into your own mind, and the dreams you actually remember, the ones from the far end of the night, are still yours to sit with, so look up whatever keeps recurring in the dream dictionary. If the vividness of those is what is on your mind, our page on why dreams get so vivid covers the other end of the same night. This guide is educational and meant for reflection, not medical advice, and it cannot diagnose anything. If someone is being hurt, if episodes are frequent, if this began for the first time in adulthood, or if it arrived alongside a medication change, speak with a qualified clinician.

10Sources

Every claim on this page that comes from somewhere else, with the somewhere else, so you can read each one in context rather than trusting a paraphrase.

• Mayo Clinic: Sleepwalking, symptoms and causes
• MedlinePlus: Sleepwalking
• National Heart, Lung, and Blood Institute: Sleep phases and stages
• NCBI Bookshelf: Physiology, sleep stages
• PubMed: Irfan, Schenck and Howell, non-REM disorders of arousal and related parasomnias, Neurotherapeutics 2021
• PubMed: Idir, Oudiette and Arnulf, sleepwalking, sleep terrors, sexsomnia and other disorders of arousal, Journal of Sleep Research 2022
• PubMed: Cataldi and colleagues, abnormal timing of slow wave synchronization processes in non-REM sleep parasomnias, Sleep 2022
• PubMed Central: Oudiette and colleagues, dreamlike mentations during sleepwalking and sleep terrors in adults, Sleep 2009
• PubMed: Castelnovo and colleagues, mental activity during episodes of sleepwalking, night terrors or confusional arousals, Nature and Science of Sleep 2021
• PubMed: Zadra, Pilon and Montplaisir, polysomnographic diagnosis of sleepwalking and the effects of sleep deprivation, Annals of Neurology 2008
• PubMed: Pilon, Montplaisir and Zadra, precipitating factors of somnambulism, Neurology 2008
• PubMed: Blanchette-Carrière and colleagues, self-reported precipitating and priming factors for somnambulism in adult sleepwalkers, Journal of Sleep Research 2026
• PubMed: Stallman and Kohler, prevalence of sleepwalking, a systematic review and meta-analysis, PLOS ONE 2016
• PubMed: Stallman, Kohler and White, medication induced sleepwalking, a systematic review, Sleep Medicine Reviews 2018
• PubMed: Petit and colleagues, childhood sleepwalking and sleep terrors, JAMA Pediatrics 2015
• PubMed: Hublin and colleagues, prevalence and genetics of sleepwalking, Neurology 1997
• PubMed: Mundt and colleagues, behavioral and psychological treatments for NREM parasomnias, Sleep Medicine 2023