Baby Nap Transitions: 4 to 3, 3 to 2, and 2 to 1
By the Growbaby team · Published Aug 7, 2026
Written against AAP, CDC and WHO guidance. How we write these guides
At a glance
- There is no official nap schedule. The AAP publishes 24-hour sleep totals and no nap-drop ages at all.
- The signal is what matters, not the calendar: a nap refused consistently, or a bedtime that falls apart.
- Usual windows: 4 to 3 naps around 3 to 4 months, 3 to 2 around 6 to 9, 2 to 1 around 12 to 18.
- Two naps to one is the hard one, and doing it at 12 months because of the number is the usual mistake.
- Give it two weeks before deciding. One bad week is illness, teeth or travel far more often than a transition.
Search for when a baby drops to one nap and you will get twelve months, fifteen months and eighteen months, from sources that all sound equally certain. It is tempting to conclude that somebody is wrong. Nobody is: there is no standard to be wrong about.
That is the useful thing to know before reading any of the numbers below, including ours. Nap transitions are a scheduling problem, not a milestone, and they are read off your own child rather than off a chart.
There is no official nap schedule
The American Academy of Pediatrics supports the American Academy of Sleep Medicine's guidelines on how much sleep children need, and those guidelines are totals across a full 24 hours: 12 to 16 hours for infants aged 4 to 12 months, 11 to 14 hours from one to two years, and 10 to 13 hours from three to five. Each of those is stated as including naps. What neither body publishes is a nap count by age, or an age at which a nap should go.
The consensus statement goes further in a direction worth knowing about. It makes no recommendation at all for babies under four months, and says why: the variation in how long and in what pattern young babies sleep is too wide, and the evidence linking it to outcomes too thin. The people best placed to publish a newborn sleep schedule looked at it and declined.
That absence explains the contradictory numbers you find everywhere else. Every age you read for a nap transition, ours included, is convention from paediatric sleep practice rather than a clinical standard, which is why they drift by months between sources without any of them being wrong.
It has a practical upshot. If your child is getting roughly the right total for their age and is not miserable, the number of naps that total is divided into is not itself a problem to solve.
What a transition actually looks like
A transition announces itself in one of two ways, and neither of them is a birthday. Either the nap is refused, consistently, by a baby who used to take it, or the naps still happen and the night pays for it: bedtime becomes a fight, settling takes an hour, or the morning starts at five for no reason anyone can name.
The test that matters is duration. Give it two weeks before you conclude anything. Illness, teething, travel, a new sibling, a house move and a developmental leap all produce exactly this picture and all of them resolve on their own, usually inside a fortnight. Restructuring the day on the strength of four bad nights is how families end up chasing a schedule that was fine.
- Refusing a nap they reliably took, most days, for two weeks or more
- Naps unchanged but bedtime collapsing, or settling stretching out
- Early waking with no illness, teeth or travel to explain it
- The last nap of the day pushing bedtime steadily later
The four transitions and their usual windows
These are the windows this site uses, and they match the wake window calculator so the two never contradict each other. Read them as where most children land, not as dates.
The first one barely registers for most families, and the sleep researchers would say that is exactly right: this is the age band they declined to make any recommendation about. Newborn sleep is not really organised into naps, so the fourth and fifth sleeps of the day fall away as the day organises itself, and the change tends to be noticed in retrospect if at all. The last one mostly happens past this site's range, somewhere between three and five years, and school is often what ends it rather than any readiness signal.
- 4 naps to 3, around 3 to 4 months: the short evening catnap goes first
- 3 naps to 2, around 6 to 9 months: often mistaken for the eight-month regression
- 2 naps to 1, around 12 to 18 months: the long, difficult one, covered below
- 1 nap to none, usually 3 to 5 years: past this site's range, and rarely tidy
Two naps to one is the hard one
The others are subtractions. This one is a restructuring, and that is why it takes weeks rather than days. Dropping from two naps to one means the stretch of awake time before the remaining nap grows by an hour or more, and there is no intermediate step available: your child either bridges the morning or does not.
The common mistake is doing it at twelve months because twelve months is a round number. The wake windows this site publishes put a one-year-old at roughly three to five hours awake between sleeps, which still accommodates two naps for a great many children. A first birthday is not a signal.
When the signals are genuinely there, moving is gentler than dropping. Push the morning nap later in fifteen to thirty minute steps across a couple of weeks until it has migrated into the middle of the day and absorbed the afternoon one. Expect an early bedtime throughout, because the day will be short on sleep while the change is happening, and an early bedtime is the pressure valve.
Expect it to be untidy. Weeks of one nap on some days and two on others is the normal shape of this transition, not a sign that it has failed or started too soon. Following the child on a given day beats holding a schedule that day cannot support.
Transition or regression?
From the outside these are the same event: a child who slept and now does not. Two things tell them apart, and neither requires a diagnosis.
A regression arrives suddenly, usually alongside a new skill or a change in circumstances, and it resolves within roughly one to three weeks with the schedule left exactly as it is. A transition builds gradually, does not resolve on its own, and only settles once the schedule is changed to match the child.
So the practical test is patience followed by arithmetic. Hold the routine steady for three weeks. If sleep has recovered, it was a regression and you were right not to rebuild the day. If it is still broken, the schedule is the thing that no longer fits, and the day needs changing rather than defending. Our guide to sleep regressions covers the ages those cluster around and why the four-month one is the odd one out.
Naps follow the same rules as nights
One thing does not vary with the schedule. Every nap follows the same safe sleep rules as the night: on the back, on a firm flat surface, in a crib or bassinet with nothing else in it. Daytime sleep is where those rules slip most, because a nap in the stroller, the car seat or on a parent's chest is convenient and feels different from bedtime.
The one worth naming is the car seat, because it is genuinely common. A baby who falls asleep in one on a journey is fine; the seat is not a place to leave a sleeping baby once you have arrived. Moving them to a flat surface is the whole of the advice.
When to raise it
Nap counts are not a clinical measure and there is no number to hit, so most of what happens here is not a doctor question. Two things are. One is total sleep sitting persistently well outside the range for your child's age, in either direction, over weeks rather than days. The other is a child who seems tired all day regardless of what the schedule does.
Anything about how your child breathes while asleep also belongs with your paediatrician rather than with a schedule. It is a different kind of question from this page's, and it is one worth asking directly.
Growing with your baby, week by week
Growbaby adapts to your baby's age: soothing that reacts to crying in real time, sleep routines for this exact stage, and Growy, an assistant that remembers your child. One subscription covers the whole family.
Frequently asked questions
Commonly somewhere between six and nine months, though the range is wide and there is no official age, because no health body publishes nap counts. The signal to watch is the third nap being refused most days for a couple of weeks, or that third nap pushing bedtime steadily later.
Usually somewhere between twelve and eighteen months, and closer to the later end than parents expect. A first birthday on its own is not a reason: the wake windows for a one-year-old still leave room for two naps for many children.
Hold the routine steady for three weeks and see. A regression resolves on its own within about one to three weeks with the schedule unchanged. A transition does not resolve until the schedule changes, because the schedule is what stopped fitting.
The small ones can pass in a week or two. Two naps to one commonly takes several weeks, and during it your child will take one nap on some days and two on others. That inconsistency is the transition, not a sign it is going badly.
No. Ages are useful for knowing what to expect and useless as instructions, since none of them come from a clinical standard. Drop a nap when your child stops taking it or when keeping it is costing you the night, not when an article says so.
Yes, and it is the usual pattern during the two-to-one change rather than a problem with it. Follow the day you are actually having, and use an early bedtime on the one-nap days to make up the difference.
Sources
This guide is for general informational purposes only. Growbaby is not a medical device and does not provide medical advice, diagnosis or treatment. Every baby develops at their own pace; for any question or concern about your baby's health or development, always consult your pediatrician or another qualified healthcare provider.