How long should it take to fall asleep?

How long does it normally take to fall asleep?

Anything up to about half an hour. The time it takes has a name, sleep latency (or sleep onset latency), and the clearest guide to what is normal comes from the National Sleep Foundation in the United States. In 2017 it published the verdict of an 18-member expert panel that had reviewed 277 studies measuring sleep objectively, then voted, over three rounds, on what signals good-quality sleep at each age (Ohayon and colleagues, 2017).

In other words: up to 30 minutes is normal from infancy to old age; more than 45 is a sign of poor sleep before 65, and more than an hour is at every age. On 31 to 45 minutes the experts could not agree — a fair reminder that there is no single minute at which a normal wait becomes a long one.

You will often read that 10 to 20 minutes is normal. That is a typical time rather than a limit: by the panel’s judgement, anything up to 30 minutes is a sign of good sleep, so 25 minutes on a given night is not a problem to solve.

Two details are easy to miss. The panel set no lower limit — under 15 minutes was simply good. And the studies behind its numbers measured sleep with recordings rather than recollection, which matters further down this page.

Does it take longer to fall asleep as you get older?

A little, and less than you might expect. A 2004 meta-analysis pooled 65 studies of 3,577 healthy people aged 5 to 102 whose sleep had been recorded through the night. In adults, the time to fall asleep did lengthen with age, but the authors called the change “very subtle”: young and middle-aged adults took about as long as each other, so did middle-aged and older ones, and the whole increase between the ages of 20 and 80 came to less than 10 minutes (Ohayon and colleagues, 2004). Between childhood and the teenage years it hardly changed.

What changes more with age is the rest of the night. In the same analysis, time spent awake after first falling asleep went up with age, and the share of time in bed spent asleep went down — the one measure that kept falling after 60. Brief waking in the night is the ordinary part of that; waking at three is normal explains why. The 2017 panel was a little more lenient at that age, too: for people of 65 and over, it could not agree that 46 to 60 minutes was poor.

Is falling asleep instantly a good sign?

Not necessarily. How quickly you fall asleep is the very thing sleep clinics use to measure sleepiness. In the multiple sleep latency test, the standard clinical measure, a person gets five chances to nap across a day, two hours apart, in a dark, quiet room, and each nap ends after 20 minutes if sleep has not come. The test, in the words of the American Academy of Sleep Medicine’s practice parameters, “is based on the premise that the degree of sleepiness is reflected by sleep latency”, and its averages shorten after sleep has been cut short or broken up (Littner and colleagues, 2005).

Healthy people tested this way took about 10 to 12 minutes on average, lying down in the dark during the day, with a wide spread from one person to the next. The same guidelines warn against judging anyone’s sleepiness from one number on its own, and this is a daytime test: it sets no cut-off for bedtime, and the National Sleep Foundation’s panel counted anything under 15 minutes as good.

But the logic carries over. If you are out within a minute or two every night, and nod off in the day whenever you sit still, it can mean you are short of sleep, or that the sleep you get is being broken up — by sleep apnoea, for example — rather than that you are especially good at it.

Why does it feel as if it takes so much longer?

Because the clock in your head is a poor judge of the wait, and it errs in one direction: people who sleep badly tend to overestimate how long they lie awake. A 2012 review of this puzzle found that many people with insomnia overestimate how long they take to fall asleep, and underestimate how long they sleep, compared with what a recording shows — a tendency the authors call “ubiquitous (although not universal)” (Harvey and Tang, 2012).

It is not that they are bad at judging time in general; that explanation was not supported. The best-supported ones were that sleep itself can be experienced as being awake, that worry inflates the estimate, and that brief awakenings add to it.

Watching the real clock makes all of this worse. In experiments that asked people to watch one as they tried to fall asleep, they worried more and took longer, and those with insomnia overestimated the wait by more than people watching a neutral display (Tang, Schmidt and Harvey, 2007).

So don’t time it. Face the clock away, keep the phone across the room, and take the numbers on this page as reassurance rather than a target: a wait that felt like an hour may well have been shorter, and checking only gives the worry something to count. What to do when you can’t sleep has the rest of the plan for a night like that.

What decides how fast you fall asleep?

Mostly two systems, and whether they agree. Sleep pressure builds the longer you are awake; the body clock makes sleep easier at some hours than at others, whatever the pressure. You drop off quickly when high pressure meets a clock that has turned towards night, and slowly when the two are out of step — after a late nap, say, or when your clock is running late, which is how people end up tired but unable to sleep. Two processes decide when you are sleepy explains the pair.

The third thing is how switched on you are when you lie down. A busy mind can hold sleep off even when both systems are ready, which is why trying harder rarely helps. How to fall asleep is the longer guide, and cognitive shuffling is one way to give a busy mind something harmless to do instead of something to stop.

Is it normal to take an hour to fall asleep?

Now and then it is no cause for concern; one slow night is not insomnia. As a pattern it is worth attention, and sleep researchers have had to decide where the pattern begins. A 2003 analysis that drew on two decades of insomnia trials, and tested possible cut-offs in a random sample, found the most defensible criteria to be taking 31 minutes or more to fall asleep — or lying awake that long after first falling asleep — on at least 3 nights a week, for at least 6 months (Lichstein and colleagues, 2003).

That is a research convention, not a diagnosis. What matters as much as the minutes is whether the nights are affecting your days.

When should you see a doctor?

If trouble falling asleep happens most nights, has gone on for months and is affecting your days, take it to a doctor. For chronic insomnia, cognitive behavioural therapy for insomnia (CBT-I) is the recommended first treatment, by the American College of Physicians (Qaseem and colleagues, 2016) and the American Academy of Sleep Medicine (Edinger and colleagues, 2021).

The other end deserves a mention too. If you fall asleep the moment you lie down and still struggle to stay awake in the day, say so: sleepiness like that has causes a doctor can look for. The techniques on this site are for an ordinary slow night, not a substitute for care.