Almost every baby sleep schedule you will find online is presented with more confidence than the evidence behind it can support. This page is the map of our sleep section, and it starts by being clear about which numbers come from a health authority, which come from cohorts describing what children happen to do, and which are conventions somebody invented. The difference matters, because it tells you which ones are worth arguing with.
Start with the 24-hour total, not the schedule
The American Academy of Sleep Medicine's consensus statement recommends 12–16 hours of sleep per 24 hours for infants aged 4–12 months, naps included, and 11–14 hours for one- to two-year-olds. It says nothing at all about how that total should be divided between night and day.[1] That is the number to judge a day against — not the nap count, and not somebody else's timings.
12–16 hours
One thing worth knowing before you read any newborn schedule: the AASM consensus covers four months and up. Below four months the task force found the evidence insufficient to make a recommendation at all.[1] So a page that gives a firm hourly target for a six-week-old is quoting something other than the consensus. At that age, patterns are genuinely irregular and vary hugely from baby to baby.[2]
Wake windows are a heuristic, not a guideline
No health authority publishes wake-window figures. They publish totals. Wake windows are a scheduling convention that parents and apps find useful — including ours — and they work well enough as a planning tool, but they are not a clinical standard and a baby who does not match the chart is not doing anything wrong. Wake window stacking explains the one part of it that reliably helps: the last window before bed should be the longest of the day. The wake window calculator will do the arithmetic from your own baby's age and times.
What the day usually looks like, by age
| Age | Naps a day | Sleep per 24 hours |
|---|---|---|
| 0–3 months | Irregular, no settled pattern | No AASM recommendation below 4 months |
| 4–6 months | 3, sometimes 4 | 12–16 hours |
| 6–9 months | 3 dropping to 2 | 12–16 hours |
| 9–12 months | 2, well established | 12–16 hours |
| 12–18 months | 2 dropping to 1 | 11–14 hours |
| 18–24 months | 1, in the afternoon | 11–14 hours |
The nap columns come from cohort data: a two-nap pattern is well established by nine to twelve months, and a single afternoon nap between fifteen and twenty-four.[3] Total sleep declines steadily across the whole period, with a wide normal range at every age.[4] If you want the developmental context alongside the sleep, our milestones section runs on the same age bands.
It is worth knowing how thin the ground under all of this is. A systematic review of napping and child development from birth to five years found 26 studies, every one of them observational and every one graded low quality.[5] Nap counts by age describe what children did. They are not trial results, and they are not targets.
The transitions, in order
- Cat napping — why naps are 30–45 minutes, and why that is usually one complete sleep cycle rather than a nap that failed.
- The 3-to-2 nap transition — usually between six and nine months.
- The 2-to-1 nap transition — usually between thirteen and eighteen.
Every one of these is easier to read from a week of data than from the last bad afternoon, and each of them temporarily makes an earlier bedtime the single most useful lever you have.
When the night is the problem
A long, cheerful wake-up in the middle of the night is a different problem from a baby who cannot settle, and it usually has a daytime cause. Split nights covers the two-hour 2am party and how the day produces it; wake window stacking covers the short final window that is the most common trigger. Fix the day before you change anything about the night.
If the schedule is right and sleep still is not
Behavioural sleep interventions are the most-studied thing in this whole field. The AASM's task force reviewed 52 treatment studies: 94% reported the interventions were efficacious, with over 80% of treated children showing clinically significant improvement maintained for three to six months.[6] Several distinct methods are supported, they differ in what they are best at, and the honest comparison — including the option of doing nothing — is in sleep training methods compared.
- The Ferber method — graduated extinction; the strongest evidence for night wakings.
- The chair method — camping out; the better fit for babies with high separation anxiety.
- Pick up, put down — popular, and with no trial evidence at all. We say so.
- Bedtime fading — for a bedtime that is simply too early.
Safe sleep comes before any schedule
Nothing on this page changes the basics: on the back, in a clear, flat, separate sleep space, for every sleep including naps.[7] Safe-sleep guidance also differs between countries in its details, more than most parents realise, so follow the version your own health service publishes rather than the loudest one online.
How to use the rest of this section
If your baby is under four months, expect irregularity and judge on feeding, weight and mood rather than on a schedule.[2] From four months, the 24-hour total is the measure that has an authority behind it, the nap pattern is a description you can work with, and wake windows are a planning convenience. Every article in this section cites its own sources and states where the evidence runs out — which, in infant sleep, is earlier and more often than the internet suggests.
