Wake Windows for a 3 Month Old

Most three-month-olds manage about an hour and a quarter to an hour and three quarters awake between sleeps, across three to five naps. This is usually the first month where the day has a shape you can recognise from one day to the next: two or three naps that turn up at roughly the same points, plus a short late one to get her to bedtime. Nothing is fixed yet, and the range of naps is wide on purpose — four sleeps one day and three the next is an ordinary week at three months.

The band and the arithmetic behind it sit on the wake windows by age overview, and the wake window calculator will add it to the time she last woke. This page is about the two things that make the third month different: the day becomes readable, and it is the last month before the one where the whole thing changes.

The month the day becomes readable

Naps at two months are wherever you can get them. At three months they start to cluster. What most parents see is a nap not long after the morning wake-up, a longer one in the middle of the day, another in the afternoon, and a short late one — twenty minutes in the pram or the carrier — that exists only to get her to bedtime without falling apart.

Two useful things follow from that. The first is that the morning nap is usually the most reliable one you have, so it is the anchor of the day rather than the one to mess with. The second is that the late catnap is the loosest, the shortest and the one that will eventually disappear. That is what a nap transition looks like from the inside: not a decision, but one sleep getting shorter and harder to get until it stops being worth chasing.

The sleep schedule generator will draw a sample day at this age off the morning wake-up, which is worth doing once for the shape rather than the times. What it shows is a plannable morning and a vague evening — and that is not a flaw in the drawing, it is what a three-month-old's day honestly is.

Write down what is happening, while it is calm

This is the argument of the whole page. Around four months a baby's sleep changes shape — the American Academy of Pediatrics puts regular sleep cycles at about four months of age, and before that the pattern is genuinely irregular. Whatever else that brings, it means the month you are in is the calmest look you will get at your own baby's rhythm for a while.

So take it. Three or four days on a printed day sheet or in a tracker — the time she woke, the time she went down, roughly how long each nap ran, what time the evening went wrong — and you will have something no table on the internet can give you: her windows rather than the average ones. When the nights change, the difference between "this is new" and "this is what she has always done" is the single most useful thing you can own, and it only exists if somebody wrote it down first.

The crying has usually turned the corner

One published fact that belongs to this month: the NHS says the amount babies cry tends to increase when they are around two weeks old, and then gradually reduces at around three months. If the evenings have been the worst part of your day since the fortnight mark, this is roughly where that eases — gradually, and with bad days still in it, but downhill rather than up. Most of what was written about the witching hour was written about the two months behind you.

False starts at bedtime

The complaint that turns up most at three months is the false start: she goes down at seven, sleeps forty minutes and wakes fully, and the evening restarts. It is almost always the last window of the day, and it goes wrong in both directions, so the fix depends on which one it was.

• The last window ran too long. Settling was a fight, she was frantic rather than sleepy, and she went down late and woke soon after. Bring the end of the last nap and bedtime closer together by about fifteen minutes.
• The last window was too short. She settled easily, which is the confusing part, and then woke cheerful forty minutes later ready for the day. She had not built up enough to hold the night. Add fifteen minutes, or push the late catnap slightly earlier so the gap to bed is a little longer.

Either way, change one thing by fifteen minutes and give it three or four days before deciding it did not work. A rebuilt day tells you nothing, because you cannot see which change did what. And if the answer turns out to be that the last catnap has become impossible to get, that is information rather than a problem — see the next section.

The fourth nap is on its way out

Somewhere around three to five months the day usually goes from four sleeps to three, and the late catnap is the one that goes. It is a band rather than a birthday, it overlaps with everything either side of it, and no health body publishes a nap-transition schedule at all — the ages on this site are compiled and hedged, which is why they are wide.

At three months this is in view rather than under way. What it looks like when it starts: the late nap takes longer to get than it lasts, or she takes it and then cannot settle at bedtime, or she simply refuses it two or three days out of five. One refused catnap is a Tuesday. Two weeks of them is a pattern. The nap transition checker is built to answer "not yet" as readily as anything else, and the 4 to 3 nap transition guide covers what to do when the pattern is real.

What four months brings

The next month is the one everything in this genre shouts about, so here is the plain version. Sleep cycles become more regular at around four months, which means more of the surfacing between cycles that adults do too and do not remember. At the same time the first published total-sleep figure finally appears: for infants from four months, twelve to sixteen hours per twenty-four including naps. Under four months, the sleep panel that publishes those numbers deliberately published none, because the normal variation is too wide — which is why every band on this page is soft.

What it is not is a clinical event, or a fortnight anybody has measured. The four month sleep changes page sets out what is actually published and what is folklore, and wake windows for a 4 month old has the next band, which is an hour and a half to a bit over two hours. Reading either of them now, in a calm month, is a better use of them than reading them at 3am.

Rolling is closer than it looks

The AAP puts rocking on the stomach, kicking and swimming with the arms at about five months, and those are the skills that add up to rolling over; by the end of that period most babies roll both ways, though the timing varies from baby to baby. Some start practising well before then, and some do it first in the cot at two in the morning.

That has one hard consequence at three months: if you swaddle, the swaddle stops at the first sign of rolling — when she starts trying, not when she succeeds. Everything else about the sleep stays as it was. Every nap and every night, babies go down on their back, on a firm flat surface, in a cot clear of blankets, pillows, bumpers and soft toys, with a fitted sheet and nothing else, in your room for at least the first six months. Nothing weighted goes on or over a sleeping baby, at any age.

The two month page covers the month behind you. Nothing here is medical advice and no page can see your baby — feeding that worries you, no weight gain, a baby who cannot be settled at all or a cry that does not sound like her usual one are all reasons to ring your health visitor or pediatrician rather than read on.