Knowledge

Data

How babies actually sleep, from ~115,000 babies · 5 million baby-days. Not a survey, not a textbook. Real logs.

All 10 datasets

The flagship chart

Total sleep by age

Night sleep barely moves. Naps do all the shrinking.

  • Night sleep
  • Daytime naps
051015369121518

Total sleep falls from 14.8 h at 3 months to 12.5 h at 18 months. Almost the entire drop is naps, 3.8 h down to 1.7 h.

What a normal night looks like

Night wakings by age

A quarter of babies still wake twice a night, at every age.

00.511.523691215181.20.3

The average falls with age, but the spread does not close: the p75 line sits at 2+ wakings throughout. Logged wakings only, so read it as a relative signal.

Dropping naps, month by month

The nap transition ribbon

The whole 4 → 3 → 2 → 1 journey, in one picture.

  • 1 nap
  • 2 naps
  • 3 naps
  • 4+ naps
0255075100369121518

Share of days at each nap count. Every column sums to 100%, so you can read a whole population mid-transition.

When babies actually go down

Bedtime by age

Bedtime hardly shifts all year: 8:15 to 8:30 pm.

6:00 pm7:00 pm8:00 pm9:00 pm10:00 pm11:00 pm3691215188:28 pm8:20 pm

The median barely moves, but families spread widely: the earliest 5% are down by 7 pm, the latest 5% closer to 10 pm.

The most stable number in baby sleep

Morning wake-up by age

Babies are alarm clocks. 7:10 to 7:30 am at every age.

5:00 am6:00 am7:00 am8:00 am9:00 am10:00 am3691215187:31 am7:13 am

Wake-up time is the single most stable figure in the dataset. Pair it with bedtime to read the whole sleep window.

How sleep consolidates

Sleep moving into the night

Daytime sleep halves as a share, 26% down to 14%.

101520253036912151824.9%13%

Consolidation is a share story, not a total-hours story. Babies move sleep into one long night rather than sleeping less.

Yes, winter is worse

Night wakings across the year

40% more night waking in January than in June.

0.450.50.550.60.65JanFebMarAprMayJunJulAugSepOctNovDec0.6330.571

Ages 4 to 12 months, both hemispheres pooled, so the winter signal is directional rather than age-adjusted. It is a strong pattern, not a controlled one.

Same sleep, very different clocks

Bedtime around the world

Norway is down at 7:46 pm. Italy at 9:08 pm.

  • Bedtime
  • Wake-up
NorwayAustriaSwedenUnited KingdomGermanyCanadaSwitzerlandColombiaUnited StatesFranceBrazilMexicoSpainChileItaly19:4621:08

Total sleep is remarkably similar everywhere (13.3 to 13.8 h). Bedtime culture is not. Later bedtimes come with correspondingly later wake-ups. Country is inferred from app geolocation.

The transitions, with their normal range

When each nap goes

4 → 3 at 6 months. 3 → 2 at 8.5. 2 → 1 at 13.3.

4 to 3 naps5.98 mo3 to 2 naps8.51 mo2 to 1 nap13.34 mo1 to 0 naps18.17 mo

Median with the typical range (p25 to p75). The ranges are wide: the 2-to-1 transition normally lands anywhere from 12 to 15 months.

The spread, not the average

How much babies differ

At 6 months the middle half nap between 2.4 h and 3.3 h.

02463691215183.76 h1.67 h

A full hour separates the middle half of babies at the same age. If your baby does not match an average, the average is the problem.

How the data was made

Only qualified days: stretches of 7+ consecutive days with a stable nap count, so these are settled routines rather than one-off days. Every figure averages within each baby first, then across babies, so every baby gets one vote.

Jun 2024 to Jun 2025 (seasonality: Jan 2025 to Apr 2026). Every number is aggregated. No individual baby or family is identifiable.

Napper analytics (Athena), aggregated