How Much Sleep Does a Baby Really Need? The Numbers Behind Infant Sleep

Parents usually hear a number when they ask how much their baby should sleep. Twelve hours. Fourteen. Sixteen.

Infant sleep is not a fixed target. Two babies of the same age can sleep for different totals and have entirely different patterns, and the same baby's sleep can change considerably from one night to the next.

So what does the research actually say?

The 12 to 16 hour range is a recommendation, not a daily target

The American Academy of Sleep Medicine recommends that infants aged 4 to 12 months sleep 12 to 16 hours per 24 hours, including naps, on a regular basis. The recommendation was developed through a formal consensus process reviewing a large body of scientific evidence.

The phrase on a regular basis is doing real work in that sentence. It does not mean every healthy baby sleeps exactly 12, 14 or 16 hours every single day.

A review of biologically normal infant sleep found substantial variation in infant sleep duration, and emphasised that individual babies do not necessarily follow the pattern of a population average.

A useful range is a reference point. It is not a score.

What does infant sleep actually look like?

One of the more informative datasets comes from the Rise & SHINE study. Researchers studied 314 infants with an average age of 6.4 months. The infants wore ankle actigraphs for seven days while their parents simultaneously recorded sleep using diaries and questionnaires.

The actigraphy data showed:

Measure	Average
Night sleep	526 minutes (8 h 46 m)
Daytime sleep	143 minutes (2 h 23 m)
Longest nighttime sleep period	460 minutes (7 h 40 m)

The averages only tell part of the story. The standard deviation for nighttime sleep was 67 minutes, and for the longest nighttime sleep period it was 100 minutes. There was considerable variation between infants.

That is one reason comparing your baby directly with another baby can be misleading.

Your baby's sleep tracker may not tell the whole story

The same study compared objective actigraphy measurements with parent-reported sleep. Parents tended to report more sleep than the actigraphs recorded.

Compared with actigraphy	Overestimate
Parent questionnaires, nighttime sleep	about 67 minutes
Parent sleep diaries, nighttime sleep	about 43 minutes
Parent questionnaires, daytime sleep	about 29 minutes
Parent sleep diaries, daytime sleep	about 31 minutes

This does not mean parents are wrong. It means different measurement methods produce different numbers. A parent may reasonably consider a baby asleep when the baby is quiet and still; an actigraphy device may classify the same period differently.

For anyone using a baby sleep tracker, the distinction matters:

A sleep number is a measurement, not an absolute biological truth.

Babies do not sleep the same way every night

A 2025 mini-scoping review examined 35 studies published between 2000 and 2024 that measured infant sleep during the first six months using actigraphy, sleep diaries, or both. It found substantial variability both between studies and within individual studies, and systematic differences between measurement methods.

A single night can be a poor representation of a baby's broader pattern. Take five consecutive nights:

Night	Nighttime sleep
Monday	7 h 10 m
Tuesday	8 h 20 m
Wednesday	6 h 55 m
Thursday	8 h 05 m
Friday	7 h 40 m
Five-night average	7 h 38 m

Looking only at Wednesday makes the night look unusually short. Looking at all five gives considerably more context. This is why data over time can be more useful than an isolated number.

Infant sleep changes as babies develop

Infant sleep does not simply become "more" or "less". Its structure changes.

Researchers studying 306 healthy full-term infants at approximately 1 and 6 months found that between those ages:

Between 1 and 6 months	Change
Night sleep	increased by 65.7 minutes
Night awakenings	decreased by 2.2 episodes
Daytime sleep	decreased by 73.3 minutes

Development changes not only how much babies sleep, but where that sleep sits in the day. A baby may gradually sleep for longer stretches at night while sleeping less during the day, and total sleep alone would show almost none of that.

Does waking during the night mean something is wrong?

Not necessarily. Night waking is common during infancy, and research shows substantial variation in both the frequency and the duration of infant awakenings. The 2025 review highlights how variable night waking can be in early infancy.

Instead of asking only "why did my baby wake up four times?", it is often more useful to ask "has my baby's sleep pattern meaningfully changed?"

That moves the question from an isolated event to a pattern.

What parents can actually look at

Rather than one number, several signals read together.

1. Total sleep

How much sleep across 24 hours? For babies aged 4 to 12 months the AASM recommendation is 12 to 16 hours including naps.

2. Night sleep

Is nighttime sleep gradually becoming longer, shorter, or staying relatively stable?

3. Longest sleep stretch

Is the longest uninterrupted period changing? This can say something different from total sleep duration.

4. Night awakenings

Are awakenings becoming more or less frequent, and is that change persistent?

5. Daytime sleep

Are naps changing at the same time that nighttime sleep is changing?

6. Consistency

Is one night unusual, or does it hold across several days?

7. Context

Feeding, routine, developmental changes and illness can coincide with changes in sleep. Coinciding is not causing. Context helps a parent read the broader pattern, not assign a reason to it.

The number is not the insight

Imagine five nights of the following nighttime sleep:

Night	Nighttime sleep
1	7 h 52 m
2	7 h 40 m
3	7 h 35 m
4	7 h 18 m
5	7 h 10 m

A traditional tracker reports the average: 7 h 27 m.

Data over time says something more useful: nighttime sleep has gradually decreased across the last five nights. The next question is what else changed, naps, feeding, bedtime, night awakenings, routine, or anything else logged alongside it.

The important distinction is between what we observed and what might explain it. A pattern can be worth noticing without being a diagnosis.

Sleep duration and safe sleep are different questions

How long a baby sleeps and how a baby sleeps safely are two separate questions.

The American Academy of Pediatrics recommends that infants be placed on their backs for every sleep and sleep on a firm, flat, non-inclined surface. The AAP also recommends room-sharing without bed-sharing, ideally for at least the first six months, and notes that room-sharing without bed-sharing can reduce SIDS risk by as much as 50 percent. The same evidence base reports that sleeping on couches or armchairs is associated with a 22 to 67-fold increased risk of infant death compared with safer sleep environments.

These considerations matter far more than whether a baby slept 13 or 14 hours on a particular day.

So, how much sleep does your baby really need?

For babies aged 4 to 12 months: 12 to 16 hours per 24 hours, including naps, is the American Academy of Sleep Medicine's recommended range.

The research says something equally important alongside it. Infant sleep is variable. It changes as babies develop. Different measurement methods produce different numbers. And one night rarely tells the whole story.

A more useful way to read infant sleep is duration, timing, awakenings, naps, change over time, and context, together.

The goal is not to make a baby hit a number. It is to understand that baby's pattern.

What Lybadora does differently

Traditional baby tracking answers "what happened?".

Lybadora goes one step further: "what is changing?", and, over time, "what might this pattern relate to, and what is worth paying attention to?"

That is the difference between recording baby data and reading it.