Measured from eyes open, not from the end of the feed
Before anything else, the definition, because a surprising proportion of “the chart does not work for us” turns out to be a measurement problem. A wake window runs from the moment your baby wakes to the moment she is asleep again. Not to the moment you put her down. Not from the end of the feed. Eyes open to eyes closed, everything in between included.
That distinction is not pedantry. A four-month-old who feeds for twenty minutes, plays for an hour, and then takes fifteen minutes to fall asleep has been awake for ninety-five minutes, not sixty. If you have been timing from the end of the feed, every window in your day is running thirty-five minutes longer than you think it is, which is roughly the margin between a nap that works and a nap that does not at that age. I spent a genuinely embarrassing number of hours reading wake window content before anyone bothered to say this plainly.
The settling time is the part people leave out most often, and it is the part that varies most. If your baby reliably takes twenty minutes to go down, the window you are actually offering is twenty minutes longer than the one you planned. Start the clock at the eyes and the arithmetic behaves.
Why an interval beats a clock time in the first year
The alternative to wake windows is what most printed schedules do: name the times. First nap at 9:00, second at 12:30, bedtime at 7:00. It works beautifully on the days your baby wakes at seven, and it collapses on every other day, which in the first year is most of them.
The reason is that sleep pressure — the biological drive to sleep that accumulates while awake and discharges while asleep — runs on elapsed time since the last sleep, not on the wall clock. A baby who woke at 5:40 has been awake for over three hours by 9:00. At four months that is nearly double what she can manage, and by the time she reaches the cot she is past the point where falling asleep is easy. The schedule did not fail because it was a bad schedule. It failed because it was anchored to a number that had nothing to do with her.
Anchoring to intervals instead makes the day self-correcting. An early morning pulls the whole day earlier; a late morning pushes it later; a short nap shortens the next window. Nothing has to be recalculated, because the rule is relative rather than absolute. This is also, quietly, the logic our calculator is built on — it solves the day backward from the sleep your baby needs rather than accumulating windows forward and letting bedtime land wherever it falls.
Somewhere in the second year the balance tips the other way. Circadian rhythm consolidates, the day compresses to a single nap, and clock times start to work again. Wake windows are a tool for the stretch when they do not — which happens to be the stretch when you most need help.
The curve is not a straight line
Awake tolerance does not creep up evenly by ten minutes a month. It moves in steps, and the steps line up with the moments when the day reorganises. Reading the table above as a smooth ramp will make some months feel like failures when they are simply flat.
The first real step comes somewhere around three to four months, when infant sleep architecture matures into something structurally closer to an adult pattern. Windows lengthen noticeably, and — infuriatingly, simultaneously — naps often get shorter, because your baby now surfaces between sleep cycles the way you do and cannot yet get herself back down. This is the machinery behind the four-month regression, and it is a development rather than a malfunction.
The second step is the nap count. Every time a nap disappears, the remaining windows have to absorb the time it used to occupy, so awake stretches jump rather than drift. Three naps to two, somewhere around six to eight months, adds most of an hour to the middle of the day almost overnight. Two to one, usually closer to fourteen to eighteen months than to twelve, is a bigger jump still. Those weeks are messy by nature, and they have a section of their own because the signals that a transition is genuinely due are nearly identical to the signals of a bad fortnight.
Within a single day the windows are also uneven. The first is typically the shortest, because the night has discharged nearly all the accumulated pressure. The last is typically the longest — published sample days put it around a third longer than the first. If you offer one number four times over, the morning nap will fight you and the evening will be chaos.
Reading her, not the timer
The chart tells you roughly when to start watching. Your baby tells you when to move. Those are different jobs, and the chart is the less important of the two.
The early cues are quiet and easy to miss: a fixed stare at nothing in particular, slower and heavier movement, a drop in babbling, ears or hair being pulled at, a face turning away from you or from a toy. The first yawn is genuinely useful, and it is usually later than people assume — a yawn frequently means you have perhaps ten minutes, not thirty. The loud signs — arching, back-arched crying, the wired flailing energy that looks like a second wind — are late. By then the window has closed and you are managing a landing rather than choosing one.
The practical method is unglamorous. Take the low end of your baby’s band, set a quiet reminder for ten minutes before it, and use that reminder to start looking rather than to start the routine. Over about a fortnight you will stop needing the reminder, because you will have learned her particular tell, and her particular tell is worth more than any published range. The chart is scaffolding you are supposed to remove.
Both mistakes look identical. The fixes are opposite.
Here is the single most expensive thing to get wrong. A baby who went down too early and a baby who went down too late produce almost the same outcome: a short nap, a difficult settle, a miserable afternoon. Parents reasonably conclude “too tired” and move the next nap earlier, which, if the real problem was too little sleep pressure, makes the following day worse. Then they conclude the schedule is broken.
The tells are in the settling, not in the nap. Sort by what happened in the cot before sleep, not by how long sleep lasted.
| What you see | Window was too short | Window was too long |
|---|---|---|
| In the cot | Content, chatting, kicking, playing. In no distress at all. | Crying quickly, arching, escalating rather than settling. |
| Time to fall asleep | Long — 25 minutes or more of cheerful non-sleeping. | Long — but frantic, with crying that builds. |
| How she wakes | Cheerful, refreshed, ready to go again. | Crying, hard to console, worse than before the nap. |
| The next window | Normal length or slightly long. | Noticeably shorter than usual — the deficit compounds. |
| What to change | Add 10–15 minutes before the next attempt. | Subtract 10–15 minutes, and bring bedtime earlier tonight. |
Adjust in ten or fifteen minute steps, and give any change three or four days before deciding it did not work. Babies do not respond to a new interval overnight, and daily tweaking is how families end up with no pattern at all. If you have overshot badly and the day has gone, an earlier bedtime absorbs the damage better than an extra late nap does — a nap after about four in the afternoon usually costs more night sleep than it returns.
What nobody selling you a chart tends to mention
I want to be straightforward about the evidence here, because this site exists partly because I could not find anyone being straightforward about it when I needed them to be.
Total sleep guidance is real, and it is properly sourced. The AASM consensus statement, endorsed by the American Academy of Pediatrics, gives 12–16 hours per 24 for infants aged four to twelve months, and the National Sleep Foundation’s recommendations cover the earlier months the AASM panel deliberately left alone. Those are panels of researchers, published in peer-reviewed journals, with their methods on the record.
Wake windows have nothing equivalent. No consensus body publishes them. Among large institutions, Cleveland Clinic is the notable one publishing an explicit chart. Raising Children Network frames the same idea as how long a baby manages before showing tired signs, without minutes attached. Many hospital sources do not quantify awake time at all. And the granular month-by-month tables that dominate search results come from commercial sleep businesses, which is not in itself disqualifying but is worth knowing.
The consequence shows up as disagreement, and it is not small. Between six and twelve months Cleveland Clinic’s upper bound runs hours longer than Raising Children Network’s. Our tables carry both ends rather than averaging them, which is why the 8–10 month row spans two hours to four and a half. That row looks unhelpfully vague. It is simply not pretending to know something it does not.
None of this makes wake windows useless. The concept — sleep pressure builds, babies tolerate less of it than adults, tolerance grows with age — is well founded. What it makes them is a heuristic rather than a measurement, and heuristics should be held loosely. If the chart says two hours and your baby is plainly finished at ninety minutes, she is right and the chart is an average of other people’s children.
When the answer is not a chart
Timing is not the explanation for everything, and there is a specific failure mode where a parent keeps adjusting windows for weeks while something else entirely is going on. A few things that are worth a pediatrician rather than another table: sleep that is deteriorating week over week rather than fluctuating; a baby whose feeding, weight gain or alertness has changed; breathing during sleep that worries you, including snoring or long pauses; and any change that arrived suddenly and has not budged. Sleep is also downstream of teeth, illness, and every new motor skill, none of which respond to better arithmetic.
For safe-sleep questions — where and how she sleeps rather than when — go to NICHD’s Safe to Sleep and the AAP directly. That is not a topic a wake window guide should be answering, and we do not try.
Every month, in detail
Each guide below covers one month specifically: the ranges, a written-out clock-time day, what tends to go wrong at that age, and a calculator already set to it. Guides publish on a rolling schedule rather than arriving all at once, so this list grows.
The newborn months
Windows measured in minutes, four to six sleeps a day, and no schedule to speak of.
- Wake Windows 1 Month Old: Why There Is No Schedule YetWake windows 1 month old babies handle: about 45-60 minutes, stretching to 75 by week four, plus why there is no real schedule at this age yet.
- Wake Windows 2 Month Old: The First Signs of RhythmWake windows 2 month old babies handle: 60-90 minutes across 4-5 naps. Read the cues, survive the evening cluster feed, and spot the first real rhythm.
- Wake Windows 3 Month Old: Is It a Schedule Yet?Wake windows 3 month old babies handle: 75-105 minutes across 4-5 naps. How to tell whether the pattern you are seeing is a real schedule or a coincidence.
The reorganisation
Sleep architecture matures, short naps arrive, and the numbers finally start holding still.
- Wake Windows 4 Month Old: The Regression-Proof GuideWake windows 4 month old babies need run 105-135 minutes. Full guide to the 4-month regression, sample day, split nights and false starts.
- Wake Windows 5 Month Old: The Short-Nap GuideWake windows 5 month old babies need run 2 to 2.5 hours. Why 45-minute naps are normal, how to read them, and when to rebuild the day instead.
- Wake Windows 6 Month Old: The Range, The Day, And Where Solids FitWake windows 6 month old babies handle: 2.25-3 hours awake, 3 naps drifting to 2, plus a sample day showing exactly where milk and solids go.
Still being written
Guides for months 7, 8, 9, 10, 11, 12 are drafted and publish on a rolling schedule rather than all at once, so each one gets edited properly. Until they land, the table above covers those ages and the calculator will build a full day for any of them.
Questions that come up constantly
- What is a wake window, exactly?
- It is the stretch of time between your baby waking from one sleep and going down for the next one, measured from eyes open to eyes closed. It includes the feed, the nappy change, the tummy time, the staring at the ceiling fan, and the fifteen minutes of settling. Parents routinely measure it from the end of the feed instead, which quietly adds twenty or thirty minutes to every window and is one of the most common reasons a chart that should fit does not.
- Are wake windows actually medically established?
- Not in the way sleep totals are. No consensus body publishes wake-window ranges. Among major institutions, Cleveland Clinic is the notable one that publishes an explicit chart; Raising Children Network describes the same idea in words rather than minutes, and most hospital sources do not quantify awake time at all. The minute-level month-by-month tables you find everywhere online originate with commercial sleep businesses. The underlying concept is sound and useful. The precision is not.
- Why do two reputable sources give completely different numbers?
- Because they are measuring slightly different things and drawing on different populations. From six to twelve months the disagreement is genuinely large: Cleveland Clinic's chart runs considerably longer than Raising Children Network's for the same age. Our tables hold both ends rather than averaging them into a tidier midpoint, which is why some ranges here look uncomfortably wide. A wide range that is honest is more useful than a narrow one that is invented.
- My baby's windows do not match her age band. Is something wrong?
- Almost certainly not. The bands describe what is common, not what is required, and healthy babies sit outside them in both directions. What matters more than the width of any single window is the shape of the whole day: is she waking cheerful, feeding well, and gaining weight? A baby who consistently needs less awake time than the chart suggests, or considerably more, is usually just herself. A sudden change in what she can tolerate is the thing worth noticing.
- Should the first window of the day be the shortest?
- Usually, yes. Overnight sleep discharges most of the accumulated sleep pressure, so the drive to sleep rebuilds from close to zero after the morning wake and the first window tends to be the shortest of the day. The last window before bed is typically the longest, often around a third longer than the first. That ramp is one of the more reliable structural facts about infant days, and it holds up much better across babies than any specific number of minutes does.
- Do wake windows still matter after twelve months?
- The idea does; the published numbers stop. Nothing authoritative covers awake time past twelve months, so any 12-to-18-month figure you read, including the one in our table, reflects commercial consensus rather than a cited source, and we label it that way. What continues to matter is the relationship between the nap, the last window, and bedtime. A toddler nap that ends too late leaves too little pressure for a reasonable bedtime, which is the mechanism behind most late-toddler bedtime battles.
Sources
- Cleveland Clinic — Sleep in your baby's first year (the wake-window chart)
- Paruthi et al., AASM pediatric sleep duration consensus (J Clin Sleep Med, 2016)
- Hirshkowitz et al., National Sleep Foundation sleep duration recommendations (Sleep Health, 2015)
- National Sleep Foundation — duration recommendations
- NICHD — Safe to Sleep
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.