It is 4:47am, and the newborn baby sleep schedule you found somewhere around midnight has been wrong for six hours. You are standing in a dark room doing arithmetic on a phone screen turned down to its lowest brightness. She fed at 1:20, or maybe 1:40 — the note you made is ambiguous. She has been awake since 4:10 and the chart said 45 minutes, so she should have gone down at 4:55, except she has been crying since 4:30. Somewhere behind the exhaustion there is a specific kind of shame: everyone else appears to have figured this out and you have not.
Here is the thing that would have helped me most at 4:47am, so I will put it before everything else. There is no schedule at this stage. Not because you are doing it wrong, but because the biological machinery that makes a schedule possible has not been built yet. A newborn's circadian rhythm — the body-clock system that ties sleep to the light-dark cycle — is not meaningfully operating at birth. It comes online across roughly the first three to four months. Until it does, sleep is distributed nearly evenly around the clock in stretches driven by hunger and by sleep pressure that builds and dissipates within an hour.
So the chart was not wrong about your baby. The chart was describing something that does not exist yet.
What follows is what the first twelve weeks actually contain, what is worth doing during them, and what is worth ignoring. Some of it is reassuring and some of it is just true. If you want the reference numbers for what comes after this stage, the full age charts on the homepage pick up where this page ends.
Need the clock times for your baby?
The calculator builds the whole day from one wake-up.
The machinery is not installed yet
Two systems govern when anyone sleeps. One is homeostatic sleep pressure, which builds the longer you are awake and discharges when you sleep. The other is circadian rhythm, which is a roughly 24-hour clock entrained by light and expressed partly through melatonin.
Newborns have the first and not the second. Sleep pressure works from birth, which is why wake windows are meaningful even in week one. But circadian rhythm develops over the first several months; melatonin production in particular ramps up gradually through this period rather than being present at birth. Without a clock, sleep has nothing to organize itself around, so it distributes itself in short blocks across all 24 hours.
This explains almost everything that confuses new parents about newborn sleep. It explains why 3am and 3pm are indistinguishable to her. It explains why the routine that worked yesterday produces nothing today — a routine is a signal, and there is no clock yet to receive it. It explains why blackout curtains in week two do very little.
It also explains why things change so noticeably around three months. The clock is not being taught by your efforts; it is developing on its own timeline, and your efforts start working when it arrives.
Forty-five to sixty minutes, and how fast that goes
A newborn's wake window — from the moment she opens her eyes to the moment she needs to be asleep again — is roughly 45 to 60 minutes in the first six weeks. Some newborns manage only 30 or 35.
That number lands badly, because a feed takes 20 to 40 minutes and a diaper change takes 5. In practice the window is: wake, feed, change, and start settling. There is no play time. There is no window where you eat lunch. Understanding this converts an experience that feels like failure into one that feels merely relentless, which is a real upgrade.
The reason it matters so much is that overtired newborns are harder to settle than tired ones, not easier. Past the window, a stress response makes falling asleep more difficult, so the baby who missed her nap does not sleep more deeply later — she fights harder and sleeps worse. This is the single most counterintuitive fact about newborn sleep and the one that costs the most sleep when it is not known.
Watch for the early cues rather than the clock: a still, glassy stare into the middle distance, red eyebrows or eyelids, jerky arm movements, turning away from your face. Yawning and crying are late cues. By the time she is crying you have missed the window and the next thirty minutes will be harder than they needed to be.
- Weeks 1-3: 30-50 minutes awake
- Weeks 4-6: 45-60 minutes
- Weeks 7-9: 60-75 minutes
- Weeks 10-12: 75-90 minutes
- Early cues: glassy stare, red brows, jerky limbs, turning away
- Late cues: yawning, fussing, crying
A newborn day, written down
This is not a newborn baby sleep schedule to follow. It is what a reasonably typical week-five day looks like when you write down what happened afterward, which is a completely different object.
The pattern to notice is that it is not a pattern. Naps run from 30 minutes to three hours with no discernible logic. The gap between feeds ranges from 90 minutes to four hours. There are five naps here; tomorrow there might be seven. The AAP notes that most newborns need to eat every two to three hours, and sleep arranges itself around that rather than the reverse.
What is worth extracting from a day like this is not the times. It is the totals: roughly 15 hours of sleep across the 24, six to eight feeds, and one longer stretch somewhere at night. Those totals are stable even though nothing else is. If your written-down day produces similar totals in a wildly different arrangement, you have a normal newborn.
| Time | What happened | Awake for |
|---|---|---|
| 6:40am | Woke, fed 25 min | - |
| 7:35am | Asleep in bassinet | 55 min |
| 9:10am | Woke, fed, changed | - |
| 10:00am | Asleep, contact nap | 50 min |
| 12:20pm | Woke, fed | - |
| 1:05pm | Asleep 35 min only | 45 min |
| 1:40pm | Woke, fussy, fed again | - |
| 2:50pm | Asleep | 70 min |
| 4:30pm | Woke, fed | - |
| 5:20pm | Catnap 25 min | 50 min |
| 5:45-9:30pm | Cluster feeding, mostly awake | - |
| 9:40pm | Asleep for the night | - |
| 1:15am | Woke, fed, back down 1:50 | 35 min |
| 4:20am | Woke, fed, back down 4:55 | 35 min |
Why the evening is the worst part
Almost every newborn has a difficult stretch somewhere between roughly 5:00pm and 10:00pm. She feeds constantly, unlatches, cries, feeds again, will not be put down, and appears to be starving despite having eaten five times in three hours.
This is cluster feeding and it is normal. Breast milk supply and infant appetite both follow daily patterns, and the evening bunching of feeds is a well-documented behavior rather than a sign of insufficient milk. It frequently coincides with the fussiest period of the day, which makes it feel like a feeding problem when it is largely a timing-of-the-day problem.
The practical implication for scheduling is that in the first eight to ten weeks, bedtime is often 9:00, 10:00 or later, and trying to install a 7:00pm bedtime during the cluster-feeding period will not work. Bedtime drifts earlier on its own as the evening cluster resolves, usually somewhere between weeks eight and fourteen.
What helps in the meantime is lowering expectations for the evening and dividing it if you have a second adult available. Trading two-hour shifts through the evening is more sustainable than both people being frayed at once. If feeding is painful, if she is not producing enough wet diapers, or if the fussiness seems extreme, that is a call to your pediatrician or a lactation consultant rather than a scheduling question.
Contact naps, and whether you are creating a problem
Almost every newborn sleeps better on a person than in a bassinet, and almost every parent has been told this will create a habit that becomes impossible to break. The worry is understandable. It is also, at this age, largely misplaced.
A newborn does not have the cognitive machinery to form the kind of durable learned association people are describing. What she has is a strong preference for warmth, motion, containment and the sound of a heartbeat — a reasonable set of preferences for a creature who spent nine months surrounded by all four. Sleeping in a still, cool, silent, flat space is the unfamiliar option, not the natural one.
So contact naps in the first eight to twelve weeks are not a training error. They are a normal newborn preference, and the babies who take them are not measurably worse at independent sleep at six months than the ones who did not. If contact napping is what gets your baby to sleep and lets you sit down, it is a legitimate choice.
Two practical caveats. The first is safety: a contact nap requires an awake, alert adult. Falling asleep holding a baby on a sofa or armchair is specifically identified as high-risk in AAP and NICHD safe-sleep guidance, and it is the scenario that catches exhausted parents. If you are at the point of nodding off, the baby goes on a safe flat surface first. The second is that it is worth also practising bassinet sleep sometimes, not because the association is dangerous but because having two routes is more robust than having one. A few attempts a day is plenty; there is no need to make every nap a project.
Should you wake her to feed?
Sometimes, and this one you should get from your pediatrician rather than from a chart, because it depends on your baby's weight and growth.
The general shape of the guidance: in the earliest weeks, many pediatricians ask parents to wake a newborn who sleeps longer than about four hours to make sure she is feeding often enough, and the AAP's newborn feeding guidance notes that most newborns need to eat every two to three hours. Once a baby has regained her birth weight and is gaining reliably, many pediatricians say you can stop waking for feeds overnight and let her sleep the longer stretch.
That threshold — regained birth weight, gaining steadily — is the one to ask about at your next appointment. It is the difference between a night where you set an alarm and a night where you do not, and it is not a decision to make from a website.
Daytime is a different matter. Capping naps to protect an overall rhythm is a tool for later. In the newborn stage, if she is sleeping during the day and feeding adequately across 24 hours, letting a long nap run is generally fine. The exception worth mentioning is that a very long late-afternoon sleep sometimes eats into the night, and gently waking from that one after two and a half or three hours is a reasonable adjustment once the pediatrician has cleared you on the weight question.
What actually works in weeks one to six
Given that a clock-based newborn baby sleep schedule is off the table, here is what is left. It is a short list, which is itself useful information, because the internet will offer you a very long one.
Morning light. Get her into bright natural light shortly after her first morning wake-up. This is the strongest available input to circadian entrainment, and while it produces no visible result in week two, it is what the developing clock will eventually use.
Dim evenings. Keep the last hour or two before her long stretch low-lit, and keep overnight feeds boring — minimal light, minimal talking, minimal eye contact. The goal is to make night feel different from day even before she can tell.
A short, identical wind-down. Three or four steps, five to ten minutes, the same every time. Not because she can learn it yet, but because you will not have the capacity to invent one at three months when it starts mattering.
A safe sleep surface, every time. Firm, flat, on her back, in her own space, nothing else in it. This is the one non-negotiable and it comes from the AAP and NICHD, not from me.
And swaddling or a sleep sack, if she tolerates it — a sleep sack was genuinely the one piece of equipment I used with my own daughter, and it was the only thing on the list of things I researched that earned its place. Swaddling has to stop the moment she shows any sign of rolling, which for many babies is somewhere around eight to twelve weeks.
- Bright light soon after morning wake-up
- Dim, quiet, minimal-interaction night feeds
- A short identical wind-down before the long stretch
- Safe sleep surface every single time, per AAP and NICHD
- Swaddle or sleep sack, stopped at the first sign of rolling
Weeks six to twelve: the ground starts moving
Somewhere in this window, most families notice the first real change. It usually shows up as a single unexpectedly long night stretch — four hours, sometimes five — that then does not repeat for a week or two.
That first long stretch is the leading edge of consolidation. Wake windows lengthen to 60, then 75, then 90 minutes. Day and night begin to differentiate: more sleep shifts into the night portion and daytime alertness genuinely lengthens, so she is awake and looking at things rather than awake and eating. The cluster-feeding evening begins to shorten and bedtime creeps earlier, from 10:00pm toward 9:00 and then 8:30.
Nap count starts falling, from five or six to four or five. None of this is on a fixed timetable and it is common for a baby to be at week-ten behavior at week seven or week-seven behavior at week eleven.
What you can start doing in this window that you could not before: defend a morning wake-up window. Pick a 30-minute band and wake her if she sleeps past the top of it. This is the first genuinely schedule-shaped action that will hold, and installing it now makes the three-to-five month period considerably easier, because the entire rest of the day hangs off the morning anchor.
Things not to spend energy on right now
Sleep training. Every major approach requires a developmental readiness that newborns do not have, and no reputable source recommends it in the first months. The AAP's guidance on getting babies to sleep addresses the newborn period in terms of routine and environment, not training. Put it down until at least four months, and later than that if you would rather.
Nap-length optimization. A 30-minute newborn nap is not a broken nap. Newborn sleep cycles are short and irregular by design. Chasing longer naps at eight weeks is a project with no achievable end state.
Schedule guilt. If a chart says a five-week-old takes four naps and yours takes seven, you have a five-week-old who takes seven naps. The chart is an average of a population, and populations do not have babies.
Most gear. The searching is genuinely the hardest part of these weeks — I lost hours to it — and the honest summary is that a safe sleep surface, something to swaddle with, and a way to make the room dark cover most of what the newborn period actually requires. Everything else can wait until you know what problem you are solving.
And finally: comparison. The variation between healthy newborns is enormous. A friend's six-week-old sleeping six-hour stretches is not evidence about your six-week-old. It is evidence about the width of the normal range, which is much wider than anyone tells you at 4:47am.
Questions parents ask
How many hours should a newborn sleep in 24 hours?
The AASM consensus recommendation for infants aged 4 to 12 months is 12 to 16 hours per 24 hours, and for the first three months most guidance describes a range of roughly 14 to 17 hours. It will not arrive in a tidy block: expect it spread across five to seven sleeps of wildly different lengths. Count across the full 24 hours rather than judging any individual stretch, and remember that these are population ranges, so a healthy newborn can sit at either end of them.
When will my newborn sleep through the night?
Longer night stretches typically begin appearing between two and four months, and the interval usually lengthens gradually rather than arriving all at once. Sleeping six to eight consecutive hours is common somewhere in the four-to-six month range, though normal variation is wide and many healthy babies continue night feeding well past six months. The first four-to-five-hour stretch usually shows up around weeks six to twelve, happens once, and then does not repeat for a while, which is normal and maddening in equal measure.
Why does my newborn only sleep when held?
Because being held provides warmth, motion, sound and containment that closely resemble the environment she spent nine months in, and the bassinet provides none of those. It is extremely common and it is not a habit you have accidentally created. What helps: warming the sleep surface briefly with your hand before transferring, waiting until she is in deeper sleep about 15 to 20 minutes in, swaddling or a sleep sack, and white noise. Whatever you use, the sleep surface itself has to stay firm, flat and clear per AAP safe-sleep guidance.
Can I put a newborn on a schedule at all?
Not a clock-based one, because circadian rhythm has not developed enough to support it. What you can install is a consistent sequence and consistent environmental signals: bright light in the morning, dim and boring nights, the same short wind-down before the long stretch, and attention to the 45-to-60-minute wake window. Those are the ingredients a schedule will later be built from. A genuine schedule generally becomes possible somewhere between three and five months, once the body clock is running.
Is it normal for a newborn to have day-night reversal?
Yes, and it is one of the most common complaints in the first month. Because circadian rhythm has not developed, sleep distributes without regard to the sun, and some newborns land on a pattern of long daytime sleeps and long night wakefulness. The correction is environmental rather than behavioral: maximize bright natural light and normal household noise during daytime wakefulness, keep night feeds dim and quiet, and avoid daytime sleeps stretching past about three hours. Most reversals resolve within a few weeks.
How long should a newborn nap be?
There is no target. Newborn naps range from 20 minutes to three hours and both ends are normal, sometimes on the same day. Short naps at this age are a function of immature sleep architecture, not of anything you did or failed to do. The number worth tracking is the 24-hour total rather than any individual nap. The one exception where nap length is worth managing is a very long late-afternoon sleep that appears to be displacing night sleep, which you can gently cap once your pediatrician has cleared you on feeding frequency.
What is a safe sleep setup for a newborn?
This is the part not to improvise. The AAP and the NICHD Safe to Sleep campaign specify: on the back for every sleep, on a firm and flat surface designed for infant sleep, in the parents' room but on a separate surface, with nothing else in the sleep space — no bumpers, pillows, blankets, positioners or soft toys. Room-sharing without bed-sharing is recommended. Please read the guidance directly at the linked sources rather than relying on a summary, and take any specific questions to your pediatrician.
When should I stop swaddling?
At the first sign your baby is attempting to roll, which for many babies falls somewhere around eight to twelve weeks but can be considerably earlier. A swaddled baby who rolls to her stomach cannot use her arms to reposition, which is why the timing matters more than the calendar. Most families transition to a sleeveless sleep sack, which keeps the warmth and the sleep-cue association while leaving the arms free. Some babies protest for a few nights; a transitional sack with one arm out is a common intermediate step.
My newborn fights sleep every single time. What am I doing wrong?
Most likely nothing, though the most common fixable cause is a wake window that has run long. Overtired newborns are harder to settle rather than easier, so a baby who has been awake for 90 minutes at four weeks will fight much harder than the same baby at 50 minutes. Try moving the settling attempt 10 to 15 minutes earlier and watching for the early cues — glassy stare, red brows, jerky limbs — instead of waiting for yawning or crying. If she seems inconsolable or in pain, that is a pediatrician call rather than a timing problem.
Sources
Know a tired parent who needs this?
Send it their way — it is free and always will be.
Keep reading
Building a full day from scratch? Start with the baby sleep schedule guide and calculator on the homepage.
Written by
Sukie Gao
Mom of a daughter, and the person behind every tool on this site.
Sukie Gao has a daughter, and when her daughter was a newborn she was exhausted and completely out of her depth. She spent months piecing together sleep advice from scattered corners of the internet, and built My Baby Sleep Schedule so the next tired mother would not have to search as hard. She writes as a parent from lived experience, not as a clinician.
Last reviewed and updated . Sukie is a parent, not a clinician — check anything that matters with your pediatrician.