A night like that can feel like a gift from nowhere, and it's tempting to give the credit to whatever came last: the cup of tea, the new pillow, the phone left in the kitchen. Yesterday's hours did most of the shaping, while you were busy with other things. From the moment you woke, a pressure to sleep was building with every hour you stayed up2. The time you got out of bed was doing its own work too. It set the hour your body clock would let that pressure have its way3. When the two meet well at bedtime4, sleep tends to come easily and run deep. The rest is mostly the ordinary give in a living body, which seldom sleeps the same way twice.
Is it normal for my nights to swing between good and poor?
How much do a good sleeper's nights really vary? Researchers pooled the records of 2,404 healthy adults in the US and Australia, leaving out anyone with signs of insomnia. Each wore a small motion sensor on the wrist at home for a week or two. Over those nights, each person's sleep typically strayed about an hour and a quarter either side of their own average, seventy-seven minutes on the sensor1, and plenty of nights went further than that. So if your nights swing, a good one here and a restless one there with no pattern you can see, an hour longer or shorter than your usual is an ordinary swing for good sleepers like these.
The youngest swung most: the 18-to-24s had the most variable sleep of any age group1, perhaps because their days run less regularly. Most of the people in the study were young, White and college-educated. If you're older, your own nights may run a little steadier. And what drives the swings? The researchers point to a mix of things around us, in our minds and in our bodies1, rather than any one cause. If you did everything you meant to and still slept badly, that fits what they saw too: poor nights come to good sleepers as well1, and there may be no one thing you could point to.
Each dot is a night: above the dotted line if it ran longer than this sleeper’s own average, below if it ran shorter. The band shows the spread researchers found in 2,404 healthy adults who wore a wrist sensor at home1: about seventy-seven minutes either side of each person’s own average on the sensor, and eighty-six by their sleep diaries. Several nights fall outside the band. Only the spread comes from the study; the order of the nights is made up.
Worked out from the figures in the studies cited
See this as a table
| Night | How long they slept that night | The usual spread, about 77 minutes either side |
|---|---|---|
| Night 1 | About 25 minutes longer | Inside it |
| Night 2 | About 1 hour 5 minutes shorter | Inside it |
| Night 3 | About 1 hour 35 minutes longer | Outside it |
| Night 4 | About 2 hours 15 minutes shorter | Outside it |
| Night 5 | About their average | Inside it |
| Night 6 | About an hour longer | Inside it |
| Night 7 | About 30 minutes shorter | Inside it |
| Night 8 | About 1 hour 40 minutes shorter | Outside it |
| Night 9 | About 45 minutes longer | Inside it |
| Night 10 | About 2 hours 15 minutes longer | Outside it |
| Night 11 | About 50 minutes shorter | Inside it |
| Night 12 | About 1 hour 20 minutes shorter | Outside it |
| Night 13 | About 10 minutes longer | Inside it |
| Last night (night 14) | About 1 hour 25 minutes longer | Outside it |
| Spread of all fourteen nights | About 77 minutes either side of their average | 8 of 14 nights inside it |
The lit night: Last night
What was building while I was awake?
From the moment you woke yesterday, something began to gather that sleep researchers call sleep pressure. They can see it in the slow, large waves of the sleeping brain. The longer someone has been awake, the more deep slow-wave activity their sleep has when it comes2, and it drains away again as they sleep through the night2. Every hour awake adds a little. A long, full day, up early and busy until evening, brings plenty of it to bed, enough for a good stretch of deep sleep.
A nap can draw on that store before the night has begun. In a lab study, healthy young men took a nap in the early evening. That night they were slower to fall asleep, with less deep sleep in its first hours5. That may be part of the story behind a restless night after a doze on the sofa that ran on into the evening. A short nap early in the afternoon is a different case, and Are naps good for you? looks at it more closely.
Why does the time I got up matter tonight?
The time you got up yesterday set the other force in your night: the body clock, which keeps its own time alongside the building pressure. You may know it as the second wind that comes in the evening, just when you'd expect to be running out of steam. Why would the clock push you awake then, of all times? It pushes hardest in the hours just before your usual bedtime67, keeping you awake while the pressure is already high, so you don't drop off in the early evening. Around your usual bedtime the push falls away. Then the two forces together give you one long, unbroken night2.
The time you get up can move that clock, probably through the daylight that meets you when you do. In one study, fourteen healthy adults spent two weeks waking at their usual early hour and two weeks waking three hours later. After the later fortnight, the evening rise of melatonin, the body's signal that night has come, arrived 2.4 hours later3. It was as though their clocks had followed their mornings. A fortnight of late mornings is more than one weekend. Still, it may help explain why a couple of lie-ins can leave Sunday night slow to come.
So when you look back over yesterday, the hours you were awake and the hour you got up are a good place to start. They aren't the whole of a night, though. These two forces make up a strong model of when we sleep and how deeply, in use for more than thirty years4. Worries at bedtime, health that day and the night before8 have their say as well.
Why did I sleep so deeply after a short night?
After a short night, the next one can come on like a tide, heavy and deep. What you're feeling is the pressure you didn't spend the night before, still there. Sleep labs have watched it happen. Researchers kept healthy volunteers awake for thirty-six hours, and when at last they slept, their sleep ran deeper early in the night2. Left to sleep as long as they liked, they slept for more than twelve hours2. That's a far longer stretch awake than an ordinary short night.
Something like it seems to reach into everyday life too. In Sweden, fifty people rated their sleep each morning for six weeks, and a short night beforehand tended to be followed by a better-rated one8. Part of what made last night feel good may have been the short one before it.
This is true of people who usually sleep well, and they're who those studies followed. When 106 people with long-term insomnia kept diaries for a month, about a third of them had poor nights that came unpredictably9. For them, a bad night didn't reliably bring a good one. If that sounds like your nights, Do I have insomnia? is worth a read.
Was it the coffee, the wine or the walk?
Perhaps yesterday evening held a glass of wine with dinner, a worry you took to bed, a walk or a coffee later than usual. Some of these show up in the night, and some hardly at all.
Alcohol helps people drop off, then breaks up the second half of the night, at every dose studied in healthy volunteers10. It shows in people's own homes too, where nights were slightly more broken after a drink within four hours of bed11. Does alcohol help you sleep? goes further into why.
The worry counts for more than it might seem. Those fifty Swedes, rating their sleep each morning for six weeks, showed something else: stress and worries at bedtime predicted a poorer night more strongly than anything else8. So if a night went badly after you went to bed worrying about something, the worry is a likelier reason than the pillow or the late screen. People differ a great deal here, though: some sleep through stressful spells largely untouched, while others sleep much worse12.
The walk seems to help a little. A single bout of exercise tends to bring slightly more deep sleep and moderately less time awake in the night13. The gains are small, but they fit the sleep that so often follows a long day outdoors. If you walked yesterday, it may have added a little to last night, on top of what the day had already built.
Coffee is the hardest to judge from the inside. In a small study at home, healthy sleepers took a large dose of caffeine six hours before bed. Their measured sleep fell by more than an hour, yet their diaries showed no clear difference14. How could anyone lose an hour and not notice? Perhaps because how a night feels the next morning can miss what happened in it. An everyday study points the other way, though. Among people having their usual evening coffee, caffeine within four hours of bed showed no link with how they slept11. How long does caffeine stay in your system? puts the two side by side.
A closer look Who was studied in these evening findings?
The everyday picture of drinking and caffeine comes from 785 African American adults, with an average age of 63, who wore wrist sensors for about a week at home11. The effects it found were small: a night only slightly more broken after a drink. It watched people's ordinary lives rather than changing them, so it shows what went together more than what caused what.
The caffeine result that points the other way came from twelve healthy sleepers at home, who took 400 milligrams of caffeine six hours before bed14, a large dose. The walk's small gains come from sixty-six studies of a single bout of exercise, pooled together13.
It’s often said
It must have been that thing I did right before bed.
Why it can feel that way
The last thing you did is the freshest in memory. When a night goes well, that's where you look for the reason, and it's the thing you want to do again.
What’s going on underneath
In the caffeine study, people's own diaries missed an hour of lost sleep that the measurements showed. The researchers suggested that not noticing may be one reason people don't follow advice about caffeine14. If we can miss a cause that large, we may sometimes credit one that did little. The evening's last cup or the new pillow may deserve less of the praise than the long day, the early start and the night before.
A closer look Does the full moon change how I sleep?
Researchers are still divided on whether it does. One team went back through sleep they had already recorded in their lab and found that around the full moon people slept about twenty minutes less and showed about thirty percent less deep-sleep activity15. It was a look back at a small group, the kind of finding that needs repeating before anyone can rely on it.
A later field study put wrist sensors on people in Toba/Qom communities in Argentina, some living with electric light and some without, and on college students in Seattle. Sleep came later and ran shorter on the nights before a full moon, most of all where there was no electric light16. Other researchers have argued against any strong effect16. With credible studies pointing both ways, the moon is still an open question, and it seems fair to leave it there for now.
What the research says
- Even good sleepers' nights typically wander an hour and a quarter or so either side of their own average, and often further. Good evidence
- Sleep pressure builds with every hour awake and drains during sleep. Well established
- After longer awake than usual, healthy sleepers sleep more deeply the next time, as seen in people and animals alike. Well established
- Waking later for a fortnight tends to move the body clock later. Good evidence
- A nap in the early evening tends to make that night's sleep slower to come and lighter at first. Good evidence
- Alcohol breaks up the second half of the night. Well established
- A big dose of caffeine late in the day can cut sleep without people noticing, though an everyday study found no link with habitual evening caffeine. Good evidence
- A single bout of exercise usually helps that night's sleep a little. Good evidence
- Studies differ on whether the full moon shortens sleep, and it isn't clear yet. Still debated
Most of these findings come from healthy sleepers, many of them young adults studied in labs. In a month of diaries from people with long-term insomnia, poor nights came unpredictably for about a third, so the recovery after a short night can't be assumed there.
Can I have more nights like last night?
Often you can, and the parts of a good day worth repeating are the plain ones: getting up at around the same time, with some morning light, and a full day awake with no long doze in the evening. A steady wake time lets the light set your clock to the same hour, and the full day brings plenty of pressure to bed with you. Late coffee is a smaller matter, though it may be worth noticing. None of it guarantees a night, because some swing from night to night comes to everyone1.
Morning is when I am awake and there is a dawn in me.
And when a night turns out ordinary, or worse, it's the same give that every sleeper's nights have. The pressure it leaves behind tends to make the next night a better one8.

