Newborns typically sleep somewhere between 14 and 17 hours over the course of a 24-hour day, though research tracking actual sleep in the first few months of life often puts the figure closer to 13 to 15 hours. That gap between guideline and reality hints at something important: “normal” newborn sleep covers a surprisingly wide range, and what matters as much as total hours is how that sleep is structured, how it shifts week by week, and what shapes it along the way.
Why Newborn Sleep Looks Nothing Like Adult Sleep
A newborn’s sleep is spread across the entire 24-hour clock in short bursts. There is no consolidated nighttime block and no real daytime wakefulness in the earliest weeks. This is not a behavioral quirk or something parents are doing wrong. It reflects the fact that the brain systems responsible for organizing sleep into a day-night pattern simply are not online yet.
The building blocks of sleep themselves change rapidly. In premature and full-term newborns, the average sleep cycle lasts roughly 46 minutes at 31 to 34 weeks of conceptional age, stretching to about 70 minutes by 35 to 36 weeks. A large portion of early sleep is classified as “indeterminate,” meaning it does not fit neatly into the categories of active or quiet sleep. That indeterminate fraction drops from about 30 percent at 31 to 34 weeks down to around 12 percent by 35 to 36 weeks, as both active sleep and quiet sleep claim a bigger share and stabilize.1PubMed. Development of sleep states in normal premature and full-term newborns In practical terms, a very young newborn cycles through sleep states quickly and unpredictably, which is why they seem to wake constantly and why stretches of deep, still sleep are so brief.
Active sleep, which resembles what adults call REM sleep, makes up a huge proportion of a newborn’s total sleep. Over the first months of life this balance gradually shifts toward more quiet sleep, and overall sleep need drops as the brain matures. Modeling of these changes suggests that infant sleep patterns are consistent with faster buildup and faster clearing of the biological pressure to sleep, combined with a circadian rhythm that is weak or essentially absent in the earliest weeks.2PLOS Computational Biology. Mapping the physiological changes in sleep regulation across infancy and young childhood
When the Internal Clock Starts Ticking
One of the most common frustrations for new parents is the feeling that a newborn has no sense of day versus night. That frustration is well-founded: the circadian system, which locks the sleep-wake cycle to a roughly 24-hour schedule, develops in stages over the first two to three months of life, not all at once.
Detailed monitoring of a human infant showed that the circadian rhythm of body temperature appeared first, becoming significant within about one week of birth. The circadian pattern of wakefulness came next, reaching significance around day 45, which coincided roughly with the period when melatonin levels began rising at sunset. The circadian rhythm of sleep itself appeared last, becoming significant only after day 56.3Oxford Academic (Sleep). The Development of Circadian Rhythms in a Human Infant Sustained bouts of wakefulness lasting 90 to 120 minutes first appeared in the second month of life, typically clustered around the time just after waking and just before falling asleep.
This timeline is reassuring because it means the erratic sleep of the first month is biologically inevitable, not a sign that something is wrong. It also means that parents who feel like things suddenly “click” around 6 to 8 weeks are noticing something real: the circadian clock is literally switching on during that window.
How Light Shapes the Developing Sleep Schedule
Because the circadian system is still forming, environmental light is one of the most powerful tools parents have for encouraging a day-night distinction. A scoping review of the evidence found that cycled lighting, meaning brighter conditions during the day and dim conditions at night, improved nighttime sleep and daytime wakefulness in infants and promoted better growth and development. Infants who got more than 100 lux of illumination during daily awake periods and experienced a stronger day-night contrast in light levels showed stronger circadian activity patterns.4PubMed Central. The role of light exposure in infant circadian rhythm establishment: A scoping review perspective
Evidence has also accumulated on potential long-term programming effects of early light exposure. Inappropriate light-dark environments during the newborn period, such as constant dim lighting in a hospital nursery or bright screens at night in a home setting, may affect the maturing visual system and the developing circadian rhythm in ways that persist beyond infancy.5PubMed Central. The Long-Term Effects of Light Exposure on Establishment of Newborn Circadian Rhythm The practical takeaway is straightforward: expose your newborn to natural daylight during the daytime and keep nighttime feedings and diaper changes as dim and quiet as possible. You are not training a habit; you are helping calibrate a clock.
Breast Milk, Formula, and the Melatonin Connection
How and what a newborn eats affects when and how long they sleep, through both mechanical and biochemical pathways. On the biochemical side, breast milk contains melatonin, the hormone most associated with sleepiness, and its concentration follows a daily rhythm. Levels are relatively low during daytime hours and rise at night, peaking around 3:00 AM at a mean concentration of about 23.5 pg/mL.6Scientific Reports. Variations in melatonin levels in preterm and term human breast milk during the first month after delivery This makes breast milk a natural source of a circadian signal, delivering a dose of melatonin that varies depending on the time of day the milk is produced.7PubMed Central. Melatonin in Human Breast Milk and Its Potential Role in Circadian Entrainment: A Nod towards Chrononutrition?
This raises an interesting question for parents who pump and store milk: a bottle expressed at 2:00 AM contains more melatonin than one expressed at 2:00 PM. Some researchers have speculated that feeding stored milk out of sync with when it was collected could theoretically send a confusing time-of-day signal, though this is still more a hypothesis than an established clinical concern.
On the question of whether breastfed or formula-fed infants sleep more overall, the evidence is mixed and probably not what many parents expect. One large study found that fully breastfed infants had longer nighttime and total sleep durations at multiple time points through the first two years of life compared to formula-fed infants, though breastfed babies also woke more frequently at night between 6 and 12 months.8PubMed. Association between breastfeeding and sleep patterns in infants and preschool children A Chinese population-based study found a somewhat different pattern: exclusively formula-fed infants had the highest total sleep percentage over 24 hours, followed by exclusively breastfed and then partially breastfed infants, though these differences weakened as infants got older.9PubMed. Feeding methods, sleep arrangement, and infant sleep patterns: a Chinese population-based study The bottom line is that feeding method affects sleep patterns in real but modest ways, and neither option guarantees a baby who sleeps through the night.
How Much Variation Is Normal
The guideline range of 14 to 17 hours already spans three full hours, and real-world variation is even wider. Trajectory analysis in a large cohort identified four distinct sleep-duration groups among young children: short sleepers (about 18 percent of the sample), those who declined to short sleep over time (about 21 percent), intermediate sleepers (about 37 percent), and long sleepers (about 24 percent).10PubMed. Trajectory of sleep duration and sleep-disordered breathing and their impact on neurodevelopment in early childhood So at any given age, some babies are naturally sleeping hours more or less than their peers, and neither end of the spectrum is automatically cause for alarm.
That said, the same study did find that children in the short-sleeper group scored about 5 points lower on cognitive development assessments at age two, and that nighttime sleep appeared to matter more for cognitive outcomes than daytime naps. This does not mean every baby who sleeps a bit less is at risk, but it does suggest that persistent, substantially short sleep over the first couple of years is something worth discussing with a pediatrician, especially if it is accompanied by other developmental concerns.
What makes this tricky is that “how much my baby sleeps” is not easy to measure accurately. A study comparing parent-reported sleep logs to objective measurement with wrist-worn activity monitors found that parents overestimated their children’s actual sleep duration by roughly one hour and 53 minutes per night. In children with habitual snoring, the overestimate was about two hours and eight minutes.11PubMed Central. Sleep estimates in children: parental versus actigraphic assessments Parents tend to record when a child was put to bed and when they got up, not the fragmented periods of actual sleep in between. This does not mean you need a lab to track your newborn’s sleep, but it is worth understanding that the number you would estimate is likely higher than reality, which in turn means that many babies who seem to be sleeping “enough” by the guidelines might actually be on the shorter end.
Swaddling and Settling Techniques
Given how fragmented newborn sleep is, parents naturally look for ways to help babies stay asleep longer. Swaddling is one of the oldest and most widely practiced techniques, and the evidence supports its effect on sleep architecture. A systematic review found that swaddling was associated with increased duration of quiet sleep and a significantly reduced number of sleep-state transitions in infants who were not previously accustomed to being swaddled.12PubMed Central. The effect of swaddling on infant sleep and arousal: A systematic review and narrative synthesis Fewer state changes means fewer opportunities for a baby to shift from deep sleep into a lighter stage and wake up.
The safety caveats around swaddling matter, though. Swaddling should be discontinued once an infant begins showing signs of rolling over, because a swaddled baby who rolls face-down cannot use their arms to push up or reposition. Most pediatric safety guidelines recommend stopping swaddling by about two months or at the first sign of rolling, whichever comes first. When done correctly during the appropriate window, swaddling appears to consolidate sleep without reducing the baby’s ability to arouse in response to a genuine threat, but the systematic review noted that evidence is still limited on arousal thresholds specifically.
Cultural Differences in What “Normal” Looks Like
Much of the guidance parents encounter about newborn sleep comes from a Western, predominantly Northern European and North American, perspective. But infant sleep habits vary enormously across cultures, and those differences extend beyond simple preferences into structurally different sleeping arrangements and expectations.
A large cross-cultural survey spanning 17 countries found that children from predominantly Asian countries had significantly later bedtimes, shorter total sleep times, and higher rates of bed-sharing compared to children from predominantly Caucasian countries. Total sleep ranged from about 11.6 hours in Japan to 13.3 hours in New Zealand. Bed-sharing with parents ranged from under 6 percent in New Zealand to over 83 percent in Vietnam. Perception of whether a child even had a “sleep problem” varied wildly, from 11 percent of parents in Thailand to 76 percent in China.13PubMed. Cross-cultural differences in infant and toddler sleep These are not small differences. A baby sleeping 11.6 hours would be flagged as concerning in some Western pediatric offices, yet it represents the population norm in Japan.
Co-sleeping practices illustrate the cultural divide especially well. In many cultures around the world, parent-child bed-sharing or room-sharing is the default and has been for centuries. In Western countries, however, families who co-sleep often face stigma or critical attitudes from healthcare providers and others around them.14PubMed Central. Sources of attitudes towards parent-child co-sleeping and their effects: A systematic scoping review Even within multicultural Western societies, bed-sharing rates differ sharply by ethnic background. A Dutch study found that bed-sharing among native Dutch mothers decreased steadily from 2 to 24 months, while Turkish and Moroccan families showed increasing rates over time and Caribbean families maintained consistently higher rates.15PubMed. Ethnic differences in prevalence and determinants of mother-child bed-sharing in early childhood
None of this means safety guidelines should be ignored. Safe sleep recommendations regarding firm mattresses, back sleeping, and keeping loose bedding away from infants are based on risk data and apply regardless of cultural context. But the broader point is that the total number of hours a newborn sleeps, the time of day that sleep happens, and where the baby sleeps during it are all profoundly shaped by culture, and what feels “normal” to one family may feel deeply wrong to another for reasons that are more sociological than medical.
When Newborn Sleep Becomes a Parental Health Issue
The fragmented nature of newborn sleep does not only affect the baby. Parental sleep quality in the postpartum period is closely linked to mental health outcomes, and the relationship runs in both directions. A study of postpartum women found that mothers with poor sleep quality had roughly three times the odds of postpartum depression compared to those sleeping well, and that association held after adjusting for a range of other factors including age, activity level, delivery type, and infant sex.16PubMed Central. Association between sleep quality and postpartum depression
The relationship is complicated further by the fact that mothers with higher depressive symptoms during pregnancy were more likely to report infant sleep difficulties at six weeks postpartum. Breastfeeding difficulties at the same time point were also associated with reported infant sleep problems.17PubMed Central. Infant Sleep Difficulties at the 6th Week and the 12th Month Postpartum: What Is their Relationship with Maternal Mental Health and Other Perinatal Factors? This means that a mother who is already struggling emotionally may perceive the baby’s sleep as more problematic, which further disrupts her own rest, which deepens the depression. The cycle is real and clinically recognized, and it underscores why asking “how many hours should my newborn sleep” is never just a question about the baby. If your newborn’s sleep is severely disrupting your own ability to function, that alone is worth raising with a healthcare provider, separate from any concern about the baby’s health.
Premature Babies and Adjusted Expectations
Sleep guidelines assume a full-term birth. Babies born prematurely follow a different developmental timeline, including for sleep. As the sleep-cycle data described earlier shows, the maturation of sleep architecture happens along a schedule pegged to conceptional age, not the date of birth. A baby born at 32 weeks will have shorter, more disorganized sleep cycles than a baby born at 39 weeks, not because something is wrong but because the brain is at an earlier stage of a normal process.1PubMed. Development of sleep states in normal premature and full-term newborns
For parents of premature infants, this means that the commonly cited 14- to 17-hour range and the expected milestones for circadian rhythm development should be adjusted based on the baby’s corrected age (gestational age at birth plus weeks since birth) rather than their chronological age. A baby born at 34 weeks who is now 6 weeks old chronologically is closer to a one-week-old full-term infant developmentally. Expecting that baby to show day-night differentiation or consolidated sleep bouts would be premature in every sense. Neonatal care teams typically frame expectations around corrected age, and parents of preemies benefit from keeping that adjusted calendar in mind well into the first year.
What the Numbers Actually Mean for You
If your newborn is sleeping somewhere in the range of 12 to 18 hours a day, is feeding well, gaining weight, and reaching developmental milestones roughly on schedule, the exact hour count matters far less than the overall pattern. The wide ranges in guidelines and research reflect genuine biological diversity, not vagueness on the part of scientists. Some healthy babies sleep 14 hours; others sleep 17. Both are fine. What warrants attention is not a number that falls slightly outside a range, but persistent patterns: a baby who seems unable to sleep more than 8 or 9 hours total despite adequate feeding, or who is excessively difficult to rouse, or whose sleep does not begin consolidating at all by three to four months.
It also helps to remember the measurement gap. If you think your newborn is sleeping 16 hours, the real number might be closer to 14 once you account for the quiet-but-awake stretches you are not noticing, especially at night. That does not mean your baby is sleep-deprived. It means the guidelines were built partly on the same kind of parent-reported estimates that tend to run high. The research world is aware of this discrepancy and treats published norms with a degree of caution you can afford to share.