Wake Windows Stopped Working? What Actually Changed

When a wake window stops working overnight, the window almost never changed — your baby did, and the number you were using went stale a few days before you noticed. In practice the culprit is one of four things: your baby needs a longer window than the chart says, they need a shorter one because they are running overtired, a nap is on its way out, or the timing was never the problem and something else is breaking the night. You can usually tell which in three nights, without buying anything. Here is how I sort it in a consult.

I make my living in this industry, so I will say the unflattering part first: wake-window charts are a starting estimate, not a prescription, and a large slice of the distress I get called about is a parent defending a number against a baby who has moved past it.

The word “overnight” is doing a lot of work

Almost nothing in infant sleep changes overnight. What changes overnight is when you notice.

The usual shape is this. A window has been working, so you stop watching closely — quite right, that is the reward for having a routine. Your baby’s tolerance drifts by ten or fifteen minutes over a week or so. For most of that week the drift is absorbed by a good bedtime and a forgiving day. Then one afternoon a nap lands short, the drift and the short nap stack, and bedtime goes badly enough that you sit up and reconsider the whole schedule. It reads as a cliff. It was a slope.

That distinction matters because it changes the fix. A cliff invites you to rip up the schedule and start again. A slope only needs the number nudged, held for three days, and left alone.

Four things that actually changed

The window got too short. The most common one, and the easiest to miss, because a baby who needs more awake time does not look tired — they look fine, then fight the nap for twenty-five minutes and sleep for thirty. Fighting the crib with energy usually means undertired.

The window got too long. The mirror image, and it looks like the first one from the doorway. An overtired baby also fights the nap, but the fight has a frantic edge to it and the day gets worse rather than flat. The tell is the second half of the night: early-morning waking and 4am parties belong to overtiredness far more often than to undertiredness. My overtired baby piece is the full separation of those two.

A nap is going. This is the change parents most often mistake for a broken window, because it breaks every window at once. The three-to-two shift usually turns up somewhere around six to nine months and often near eight; two-to-one is generally a thirteen-to-eighteen-month event and twelve months is frequently too early. Betteroo’s nap-transition walkthrough sets out the readiness signals for each one, which is the part worth reading before you assume you are in a transition — most families who think they are dropping a nap are actually two weeks early.

Nothing about the schedule changed. Teeth, a cold, a new skill practised at 5am, a house that got warmer as the season turned, a return to work that moved bedtime by forty minutes. Timing gets blamed because timing is the thing you can adjust from your phone.

The three-night test that settles it

Do not rebuild the schedule. Run this instead, and change exactly one thing.

Pick the direction you suspect and move the last wake window before bedtime by fifteen minutes — longer if you suspect undertired, shorter if you suspect overtired. Hold it for three nights, because one night tells you nothing and five nights is you avoiding a decision. Keep everything else identical: same room, same routine, same order of operations.

Then read it on two measures only. How long did settling take, and when did the first long stretch end? Too short shows up as faster settling within two nights. Too long shows up as a later first waking. If neither moves, that is an answer too: it is not the window, and you can go look at the room and the health picture instead.

Three nights of notes, then act.

Where the apps help and where they don’t

A tracker is genuinely good at the thing you are bad at right now, which is remembering what yesterday actually looked like rather than what it felt like. Predictive features — the nap-window forecasters in the better apps — are honest pattern inference and they beat most parents’ guesses once they have a fortnight of clean data.

What no app can do is tell you why. A prediction says when the next nap is likeliest to take; it does not know your baby is cutting a molar, and it cannot separate an undertired fight from an overtired one, because both look like a failed nap in the log. My apps triage sorts which tool answers which question, and the short version is that trackers describe, plans instruct, and neither diagnoses.

Use the log to check your own memory. Do not outsource the judgement call to it.

When it isn’t the schedule at all

Two floors are worth knowing. Regular sleep cycles do not really establish until around four months — the AAP is explicit that babies do not have regular sleep cycles until about that point — so a newborn whose windows “stopped working” was never running a schedule to begin with. And brief night waking does not stop being normal just because you have a plan; a six-month-old surfacing and resettling after a few minutes is the expected pattern, not a failure of the day’s timings.

The other floor is medical, and it is not fine print. New pain, changed breathing, feeding that has altered, weight worries, a baby who is suddenly inconsolable rather than cross — none of that is a wake-window problem and none of it should be run through a schedule adjustment first. That is a pediatrician conversation, full stop.

What I’d tell a client

Stop reading charts for a week. Pick the direction you suspect, move the last window by fifteen minutes, and hold it three nights while changing nothing else. If settling gets faster, you were undertired; if the first waking moves later, you were overtired; if neither budges, the schedule is not your problem and more schedule work will not find it. And if you are on your fourth adjustment in ten days, the adjusting has become the problem — that is the same failure mode as sleep training that isn’t working, and the fix is the same one: pick something reasonable and hold it long enough to be wrong properly.

Not sure which of the four changes you're in? Take the free quiz A few minutes, no card. It asks about age, naps and the last few nights, then puts the wake windows inside the rest of the day. Not a medical assessment — symptoms still go to your pediatrician. See what it says →

FAQ: wake windows that stopped working

Why did my baby’s wake windows suddenly stop working?

Usually because their tolerance drifted over a week or two and you noticed on the day a short nap made the drift visible. Less often it is a nap transition beginning, and sometimes it is not the schedule at all — illness, teeth, a new skill or a warmer room will all break a working day.

Should I lengthen or shorten the wake window?

Test rather than guess. A baby fighting the crib with energy and then sleeping briefly is more often undertired; a frantic fight plus early-morning waking points to overtired. Move the last window fifteen minutes in the direction you suspect and hold it three nights.

How long should I give a schedule change?

Three nights to read a single fifteen-minute change, and about two weeks before judging a whole new schedule. Changing something every night guarantees you learn nothing, which is the most common reason families conclude that “nothing works.”

Is waking at the same time every night a wake-window problem?

Sometimes, but a fixed clock time points more often at the room, the sleep cycle structure or a habit than at the day’s timings — I separate those cases in waking at the same time every night.

Do I need an app to fix this?

No. A free tracker or a piece of paper is enough to run the three-night test. An app is useful for remembering the fortnight accurately; it will not tell you which of the four changes you are in.