← the guide

When the nap is short

4-7m · 7-12m · 12-24m · 2y+ · Updated 2026-08-25

A twenty-five-minute nap is not a failed nap. It is a partial one, and the difference matters, because it tells you exactly what to do next: the sleep that happened drained real pressure, so the next stretch of awake time will run shorter than usual. That is the whole practical consequence. Not a ruined day — a shorter day.

The asymmetry worth knowing

Sleep pressure is a quantity, not a switch. Twenty-five minutes of sleep removes roughly twenty-five minutes' worth of it. Alexis Dubief's image is a balloon that fills while your child is awake and deflates when they sleep — and, crucially, "the balloon also deflates easily: fall asleep for 10 minutes on the ride home and the balloon deflates."

Both books also draw a line at about thirty minutes for a restorative nap. Weissbluth: "consider a sleep period a restorative nap if it is about an hour or longer. Forty to forty-five minutes is sometimes enough… sleep periods shorter than thirty minutes should not count."

Read those two facts together and you get the asymmetry that governs a short-nap day: a brief sleep releases pressure without delivering restoration. It buys back time on the clock but not much on the batteries. This is why a five-minute doze in the car at four in the afternoon can cost you the evening — it flattened the drive that bedtime was counting on.

What to do, depending on where the nap fell

If it was an early nap and more sleep is coming today, shorten the next window. Weissbluth's own adjustment is about twenty minutes earlier than the usual interval — no more than that, because too little accrued pressure is the other way a nap fails. Watch the child rather than the clock; if they seem cheerful and steady, don't force a re-do.

If it was the last nap of the day, move bedtime earlier — thirty to sixty minutes, tonight. The final window of the day is normally the longest one a child is asked to hold, about one and a third to one and a half times any other, and a short last nap makes that stretch impossible on the pressure that's left. Dubief's rule for this is direct: an earlier bedtime after a bad-nap day.

In both cases, resist the late bridging catnap. Weissbluth is blunt about the trade: "it is better to have no nap and an early bedtime than a late nap and a late bedtime." A 4 p.m. drowse deflates the balloon and delays bedtime at the same time, which is the worst combination available.

If you want to try for more sleep straight away, the books cap the effort rather than the outcome: about twenty minutes of resettling for a baby, up to an hour of quiet opportunity in the cot for a toddler. Then let it go and aim at the next window.

Short naps by age

Short naps are not one phenomenon. Before about six months, forty-to-fifty-minute naps are simply what naps are; they consolidate toward the ninety-minute shape somewhere between six and nine months. A habitual thirty-five-minute nap past six months is a fixable habit rather than a verdict — the books' two moves are waking the child gently five to ten minutes before the predictable end for five to seven days, or leaving them in the dark for fifteen to thirty minutes after a micronap (first and second naps only, never the late one).

And for a toddler on a single nap, a short nap is the whole day's daytime sleep. There is no second chance later in the day to absorb it, which is precisely why the bedtime lever exists.

What a short nap actually costs, measured

Here is my own data, 1,531 logged naps of one child, sorted by age band — the share of naps that ended under forty minutes:

The short nap is not an occasional accident that happens to you; between four and seven months it was more than a third of all naps, and the median nap in that band was a short one. Then it faded on its own. Nobody fixed it. What changed was the child's ability to link sleep cycles — the consolidation both books put at six to nine months, arriving on schedule in a single real dataset.

Two honest caveats: this is one child, and short naps in the second year are rare enough (5%) that when one happens then, it is worth a glance at the day around it rather than a shrug.

Why one short nap is not a verdict

Sleep latency — how long it takes to fall asleep — sits at around nineteen minutes across the whole 0–2 year range, and it is "strongly related to the homeostatic sleep pressure that has built since the last sleep period." When a nap starts too early on the curve, the child lies there. When it starts too late, the second wind has already arrived and they lie there too. A short nap is usually one of those two, and both are timing, not character. Individual variation dominates anyway: whether a child is a long or short napper is a stable, innate trait from about six months to about twenty-one months.

What NaniNani does with it

The tide already knows. A short nap shows as a partial drain — the water drops and starts rising again from a higher point than a full nap would leave — and every forecast after it is computed from where the curve actually is, so you don't have to do the arithmetic yourself. For the first three-quarters of an hour after a short nap, the app says so out loud, then returns to normal language, because a state that stays on screen all afternoon stops being information.

One honest limit: the model treats twenty-five minutes of nap and twenty-five minutes of car-seat doze as the same pressure release, because in terms of Process S they largely are. It cannot see that motion sleep is lighter and less restorative. So the percentage will be honest about pressure and silent about quality — your read of the child covers the second part, as it does everywhere else in this app.

Where this comes from