Part Two / The First Year
Organize Your Baby
Sleep rhythms, nap transitions, wake times, and normal variation.
In this note
The one thing to remember
Infant sleep organizes itself in a broadly predictable sequence, but how much a child sleeps and when they make each transition vary enormously. The books give confident schedules that contradict each other. Use the ranges as orientation, and anchor your day on wake time, which is the one thing that doesn't vary much.
Cutting through the noise#
Because the experts flatly disagree, and neither says so.
Take the two dominant sleep books and ask them about a six-month-old:
| Total sleep | At night | Naps | |
|---|---|---|---|
| Ferber (Solve Your Child's Sleep Problems) | ~13 hours | 9.25 hours | Two, 1–2 hours each |
| Weissbluth (Healthy Sleep Habits, Happy Child) | ~14 hours | 12 hours | Two, ~1 hour each |
That's a three-hour difference in recommended night sleep for the same baby. And Weissbluth doesn't just differ — he warns that a child sleeping only nine hours at night is a serious problem, describing under-slept children as "socially demanding, bratty, and fussy," behaving "somewhat like hyperactive children," and later becoming "fat kids."
Nine hours at night is exactly what Ferber recommends.
So one book's target is the other book's crisis. If you've read both and felt anxious, the anxiety came from the contradiction, not from your child.
Where the disagreement comes from: both are describing population averages, and averages hide enormous individual variation. Neither number is wrong; both are presented with a confidence the data doesn't support.
The 60-second recap#
- Newborn sleep is fragmented. A longer night stretch starts emerging around two months.
- Naps consolidate: many irregular ones, then three, then two, then one, then none.
- Variation between healthy children is huge — larger than most schedules acknowledge.
- Wake-up time is the consistent thing. Bedtime is what varies.
- Keeping a child up later does not produce a proportionally later morning.
- If a child needs more sleep, move bedtime earlier. You can't move the morning.
- A schedule can make family life workable. It can't override biology.
What to actually do#
RECOMMENDED - ModerateAnchor the day on wake time and the child's actual tiredness cues
Wake time is the stable variable across children. Building from it is more reliable than building from a book's bedtime
RECOMMENDED - ModerateIf night sleep seems short, try an earlier bedtime
Because morning wake-up barely moves, earlier bedtime is the only lever that actually adds sleep
OPTION - ModerateIntroduce a schedule gradually as sleep naturally consolidates
Fine to encourage structure. The strict "impose a schedule from day one" school isn't supported
SKIP - HighSkip comparisons with another child or a book's exact timetable
The normal range is so wide that most comparisons are meaningless
SKIP - ModerateDon't expect a later bedtime to buy you a later morning
It usually just buys you a shorter night and an overtired child
What the data actually shows#
The averages#
From a meta-analysis of sleep-duration studies:
| Age | Pattern |
|---|---|
| Around 2 months | Big jump in the longest sleep period — night sleep consolidating |
| Around 4 months | Roughly three naps |
| 9–10 months | Average is two naps |
| 18–23 months | Average moves to one nap |
| Around 3 years | Many children drop the last nap |
Total sleep runs about 16 hours a day for newborns, falling to 13–14 hours around age one.
The variation, which matters more#
A 2016 study in the Journal of Sleep Research used data from a sleep-tracking app: 841 children, 156,989 recorded sleep sessions. (That's roughly 200 sessions per parent. Take a moment for those parents.)
The size of the dataset lets you see the spread, not just the middle. For night sleep at six months:
| Measure | Hours |
|---|---|
| Average | 10 |
| 25th percentile | 9 |
| 75th percentile | 11 |
| Full observed range | 6 to 15 |
A six-month-old sleeping six hours at night and one sleeping fifteen are both in the data. This is why the books disagree — and why matching a published schedule is not a goal your child can succeed or fail at.
Daytime sleep shows the same picture: the longest nap grows from about an hour to about two hours over the first two years, with some children barely napping and others going three hours at a stretch. The two-nap-to-one-nap transition is similarly spread out — most children have two distinct naps around eleven months and one by nineteen or twenty months, with a long messy middle.
The exception: wake-up time#
Here the variation collapses.
- At 5–6 months, most children wake between 6:00 and 8:00 a.m.
- By age 2, the window narrows to about 6:30–7:30 a.m.
Put those two facts together — night sleep varies by up to nine hours, wake time varies by about one — and the arithmetic does itself. Bedtimes must vary a lot. If your child needs more sleep, that has to come from the front end. Trying to engineer a later morning by pushing bedtime back will generally fail.
Decision guide#
| Situation | Worth trying | Don't assume |
|---|---|---|
| Mornings start too early | Move bedtime earlier and protect enough daytime sleep | That a later bedtime will produce a later wake-up |
| Nap transition is messy | Adjust gradually; watch mood and night sleep over several days | That one bad day proves the nap should go |
| Your child's schedule differs from a friend's | Look at your child's total sleep, mood, and functioning | That average equals required |
| You need predictability for work or older kids | Build consistent wake, feed, wind-down, and bedtime cues | That every day will follow it |
| Your child seems constantly overtired | Test an earlier bedtime for several days before concluding anything | That more awake time will make them sleep harder |
On imposing a schedule from day one#
Oster was earnestly told by a fellow economist that it's essential to get a baby on a schedule immediately on arriving home — decide when they eat and sleep and impose it. Babies love it, apparently. A whole category of books takes this line.
She didn't do it. The evidence she found for the strict version is thin, and her own experience is instructive: Penelope ended up on a completely ordinary schedule anyway — night sleep, three naps, then two, then one, then none — same as everyone else's child. What was hard wasn't establishing the schedule; it was knowing when each transition was due. There's no test for readiness. Their nanny once left the room during lunch and came back to find Penelope asleep in her food.
Why the strict-schedule appeal is understandable: it promises predictability at the moment you have none, including predictability about when you can sleep. That's a real benefit. Just don't confuse the appeal with evidence.
Why the sleep question matters beyond convenience: too much daytime sleep and bedtime becomes impossible. Too little and the child is overtired and bedtime becomes impossible. Both failure modes end the same way, which is why finding the actual amount your child needs is worth the observation time.
Numbers in context#
| Number | What it means |
|---|---|
| Night sleep at 6 months: 6–15 hours observed | The single most useful number here. If your child is anywhere in this range, they are somewhere real children are |
| Wake window: 6:00–8:00 a.m. at 5–6 months | Your fixed point. Build the day backwards from it |
| Ferber vs. Weissbluth: 3 hours apart on night sleep | When two authorities differ by that much, neither is describing a rule |
| Nap transitions: 9–10 months to two, 18–23 months to one | Population guideposts. Individual children move months either side |
Practical playbook#
- Observe before you intervene. Watch the natural pattern for several days. You need a baseline to know whether a change worked.
- Anchor the morning and the bedtime routine. These are the two fixed points; everything else can float.
- Change one thing at a time, in small increments. Otherwise you learn nothing from the result.
- Judge by sleep, mood, and family function — not by how tidy the schedule looks.
- Expect disruption from illness, travel, developmental leaps, and nap transitions. A bad week is a bad week, not a regression that needs a strategy.
Where the evidence runs out#
- Whether any particular schedule causally improves developmental outcomes.
- The right total sleep amount for an individual child. The ranges are wide and there's no test.
- Any reliable way to predict when a child should drop a nap.
Our family's notes#
- Current wake window:
- Current nap pattern:
- Current bedtime and wind-down routine:
- Signs this child is under-slept (specific to them):
- Next single small experiment, and how long we'll run it: