Everything we know about
how babies actually sleep.
Wake windows, nap transitions, how much sleep is normal at every age — the reference we built Cradle on, in plain language. Free to read, free to share, no account needed.
Everything on this page, written out properly — what the numbers mean, what to do with them, and what we will not claim from them.
They are not all the same length. The first stretch after morning wake-up is the shortest of the day; the one before bed is the longest. Repeating a single number is the most common way a schedule quietly goes wrong.
Anything here worth passing on, tap Share and it goes straight to your camera roll or messages, already sized for wherever you're posting it. Yours to use — no credit needed.
Sleep logs by time of day
n = 2,146 sleepsNight waking is the norm, not the failure
299 of 2,146 sleeps were recorded in that window. Not planned, not scheduled — typed in the dark, one-handed, by somebody who had been awake for twenty minutes and would be awake for another forty.
It's the reason the app is dark by default, the reason the sleep timer is a single tap from the home screen, and the reason the Live Activity sits on the lock screen so you never have to open anything at all.
What parents actually log
share of all entriesFeeds and sleeps run almost dead even — 2,154 against 2,146. Nappies follow at 1,201. Pumping and medication are the long tail, and that's exactly what you'd expect: they matter enormously to the families who need them and not at all to everyone else.
Read: what this data can tell youLogged live, not from memory
n = 2,146 sleeps1,352 of 2,146 sleeps were written down inside sixty seconds of starting. A live timer produces that shape by construction — but it also means most of this data is contemporaneous rather than reconstructed at the end of the day.
Read: how much sleep is normalMedian logged wake window by age
0–5 monthsRead this one carefully
Cradle suggests the window, the parent logs the window, and then we count the match. A wrong ladder would produce data that looks just as tidy. So this chart shows what families recorded — it is not evidence that the schedule is correct, and we don't present it as any.
The ladder comes from published pediatric sleep research, not from this chart. That's the source — the chart is only what families recorded.
Read: why the same schedule stops workingHalf of it happens before five months
6 transitions, newborn → preschoolWhy the early months feel relentless
The gap between one drop and the next roughly doubles each time — six weeks, then ten, then fourteen, twenty-four, and finally ninety-six. The same amount of change is packed into the months when you have the least capacity for it.
Which means a schedule that stops fitting is the ordinary case early on, not a warning sign. Something genuinely changes every six to fourteen weeks until about five months.
Read: why the first five months feel relentlessAggregate counts from Cradle's own production logs, 11 May – 14 August 2026, re-derived on 14 August 2026. Totals are totals across all logs in that window.
Every figure describes what parents recorded in the app. Cradle is not a monitor and does not observe your baby. A missed log is invisible to us, so these are floors, not ground truth.
Nothing here says Cradle causes babies to sleep longer, or better, or sooner. We have not run a controlled trial. Any such claim would need one, and we won't make it until we have.
These describe everything logged, not a typical household. Logging is uneven — some families record everything, many record once. Don't read any number here as "what a Cradle baby does".