Newborn Sleep Week by Week: What to Expect in the First 4 Weeks

You've been home for four days. Nobody sleeps for more than two hours at a time. Your baby seems to sleep all day and be wide awake at 2am. Yesterday they slept 19 hours. Today they've been up since noon and it's now 4am. You're trying to figure out if any of this is normal — and if it will ever follow any kind of pattern. Here's the honest, week-by-week answer.

 

Newborn sleep in the first four weeks is biologically chaotic — not because something is wrong, but because the brain does not yet have the tools to sleep any other way. What follows is a week-by-week breakdown with precise numbers, what is normal at each stage, the warning signs that justify a call to your pediatrician, and when patterns begin to emerge.

 

For the broader picture of newborn sleep from birth to 3 months, the full 0-3 month sleep overview covers the complete developmental arc. And if your baby only sleeps when held, this guide on babies who only sleep on you addresses exactly that.

 

The Quick Reference — Newborn Sleep Hours by Week

 

Total sleep requirements in the neonatal period range from 14 to 18 hours per 24 hours, distributed across multiple sleep periods, with gradual consolidation beginning around weeks 3 to 4 as the suprachiasmatic nucleus begins responding to environmental light cues (Rivkees, Pediatrics, 2003).

 

Week	Total sleep	Sleep period	Wake periods	Distinctive sign
Week 1	16–18h	1–3h	8–12×	High active (REM) sleep — frequent Moro startle
Week 2	16–18h	2–3h	7–10×	First slight day-night differentiation begins
Week 3	15–17h	2–3h	6–9×	First slightly longer nighttime stretch possible
Week 4	14–17h	2–4h	5–8×	More recognizable patterns starting to emerge

 

Week 1 — Understanding the First 7 Days

What's happening biologically

A week-1 newborn spends approximately 50% of sleep time in active REM sleep — visible as rapid eye movement under closed lids, irregular breathing, occasional smiles or small body movements. This is not poor-quality sleep. Research on neonatal sleep documents that the motor activations during active sleep contribute directly to sensorimotor map development and neural synapse formation — the brain is building its connections during this apparent restlessness (Graven & Browne, Newborn and Infant Nursing Reviews, 2008).

 

The other defining feature of week-1 sleep is the Moro reflex — the startle response that causes the baby to fling their arms outward. This triggers at every sleep-cycle transition (every 45 to 50 minutes) and routinely wakes babies who are not swaddled. It is not anxiety or discomfort — it is an involuntary brainstem reflex that diminishes gradually between 3 and 6 months.

 

What's normal in week 1

Total daily sleep of 16 to 18 hours, across 8 to 12 periods of 1 to 3 hours each. No day-night differentiation — the circadian clock does not yet exist. Feeding every 2 to 3 hours is biologically necessary, determined by the newborn's gastric capacity of approximately 7ml at birth. The absence of any predictable schedule is not a problem to solve. It is an accurate description of a neurologically normal week-1 newborn.

 

Warning signs in week 1

A newborn impossible to rouse for feeding after 4 consecutive hours — especially combined with visible yellowing of skin or eyes, or fewer than 6 wet diapers per day after day 4 — requires an immediate pediatric call. Jaundice-related sleepiness is clinically distinct from normal newborn sleep: it reflects inadequate caloric intake and elevated bilirubin levels, both requiring prompt evaluation (AAP Neonatal Jaundice Clinical Practice Guideline, Pediatrics, 2022).

 

For strategies to help your newborn sleep off you once the first week settles, the bassinet transition guide covers the practical protocol.

 

Week 2 — The First Subtle Shifts

What changes

The most significant mechanical change in week 2 is gastric capacity. The newborn's stomach has expanded from approximately 7ml at birth to 22 to 27ml by end of week 2. Research on neonatal gastric development documents this tripling of capacity and its direct effect on feeding interval length — slightly longer intervals between feeds translate into slightly longer potential sleep periods (Zangen, Ciarla, Zangen, Di Lorenzo, Flores, Cocjin & Reddy, Archives of Disease in Childhood, 2001).

 

A very tentative day-night differentiation may begin to emerge — not a pattern, but a tendency for the longest sleep period to occur slightly more often at night. This is not yet reliable or consistent.

 

What parents often mistake for a problem

Cluster feeding in the late afternoon and evening — the baby who feeds every 45 to 60 minutes between 5pm and 10pm — is a normal week-2 behavior frequently interpreted as insufficient milk supply. It is in reality a regulatory mechanism: the baby is front-loading caloric intake before the longest potential overnight stretch and simultaneously driving milk production. Intervening to space out feeds during cluster feeding is counterproductive and contrary to feeding guidelines before 6 weeks.

 

What's not normal in week 2

Weight loss exceeding 10% of birth weight without a recovery trajectory by day 10, or fewer than 8 feeds per 24 hours. Both warrant a same-day pediatric call.

 

Week 3 — The First Glimpse of Pattern

What changes

Week 3 frequently brings the first major developmental disruption. Paradoxically, sleep may appear to worsen before improving: the baby is more wakeful, harder to settle, and more demanding of contact. This is a normal and temporary neurological event — not a behavioral problem, and not caused by anything the parent has done.

 

Toward the end of week 3, some babies produce their first slightly longer nighttime stretch — 3 to 4 hours consecutively in favorable cases. This is not yet consistent, but it represents the first signal that the neurological groundwork for longer sleep is beginning.

 

The day-night confusion peak

Day-night confusion often peaks in week 3 — the baby is most alert between 11pm and 3am. Consistent morning light exposure — getting the baby into natural daylight within the first 30 minutes of the morning wake period — begins calibrating the biological clock even at this age. Research on circadian entrainment in infants documents that morning light activates retinal photoreceptors and starts influencing the suprachiasmatic nucleus timing, significantly accelerating day-night differentiation compared to constant dim light environments (Tsai, Barnard, Lentz, Thomas & Landis, Pediatrics, 2012).

 

What to do — and not do — in week 3

Introduce a short, repeatable pre-sleep sequence for the evening: bath, feed, darkness, white noise in a fixed order. This is not sleep training — it is the beginning of the conditioned cue sequence that will become a powerful sleep signal over the following weeks. What not to do: deliberately spacing out feeds to increase tiredness. At 3 weeks, this delays caloric intake without extending sleep and is counterproductive at this age.

 

Week 4 — The Last Week of the Fourth Trimester

What changes

Week 4 is often the first week parents notice the beginning of regularity — not a schedule, but patterns that begin to repeat. The longest overnight sleep period may extend to 4 to 5 hours for some babies. Sleep cues become slightly more readable: eye rubbing, staring into space, reduced responsiveness, yawning. Responding to these cues within 10 to 15 minutes of their appearance prevents the overtired state that makes settling significantly harder.

 

The 5-6 week mark — what to expect next

Immediately after the first hints of improvement in week 4 comes one of the most disorienting phases of the first three months: the peak crying and fussing period around weeks 5 to 6. This is the Period of PURPLE Crying — a well-documented developmental phenomenon in which inconsolable crying peaks, particularly in the evening, and resolves naturally around 3 to 4 months. Population data confirms this crying peak at 6 weeks across cultures (Barr, Barr, Fujiwara, Conway, Catherine & Brant, CMAJ, 2009). Knowing it is coming prevents parents from interpreting a normal developmental event as evidence that something has gone wrong.

 

What parents can do in week 4

Begin distinguishing the different cry types — hunger, fatigue, and discomfort have distinct vocal patterns that most parents start to recognize around this time. If breastfeeding is well established, a dummy (pacifier) can be introduced — the sucking reflex has a documented calming effect and pacifier use is associated with reduced SIDS risk. Maintain a consistent sleep environment: complete darkness for nighttime sleep, white noise at 65 to 70 decibels, room temperature 18 to 20°C.

 

For what comes after the newborn phase — the significant architectural shift that happens at 4 months — the 4-month sleep regression guide explains what to expect next.

 

The 3 Signs That Need a Pediatric Call — At Any Week

These apply across all four weeks. If any are present, call your pediatrician the same day.

 

1.      Impossible to rouse for feeding after 4 consecutive hours — especially combined with fewer than 6 wet diapers per day after day 4, or any yellowing of skin or eyes. This combination suggests possible jaundice-related sleepiness or dehydration.

2.     Continued weight loss after day 5 with no recovery trajectory — the baby should regain birth weight by day 14. Weight still declining after day 5 without improvement requires evaluation.

3.     Cry absent or unusually weak — a very quiet, low-effort newborn can be more clinically significant than one who cries loudly. A weak or absent cry in a week-1 or week-2 newborn warrants same-day assessment.

 

What Actually Helps Sleep in the First 4 Weeks — Evidence-Based Only

Five approaches with evidence behind them. Each targets a specific biological mechanism.

 

1.      Continuous white noise (65–70 dB). Replicates intra-uterine sound levels, masking the sudden noise changes that trigger the Moro reflex. Maintain throughout the entire sleep period — not just at sleep onset.

2.     Swaddling (until rolling signs appear). Reduces the amplitude of the Moro reflex and the frequency of startle-induced waking. Research documents a 28% increase in uninterrupted sleep duration with swaddling — discontinue immediately when rolling attempts begin (van Sleuwen, Engelberts, Boere-Boonekamp, Kuis, Schulpen & L'Hoir, Pediatrics, 2007).

3.     Complete darkness for nighttime sleep. The most powerful circadian signal available in weeks 1 to 4. Pairing darkness with nighttime interactions and daylight with morning interactions creates the environmental contrast that begins supporting the developing circadian rhythm.

4.     Careful transfer technique. Precondition the sleep surface with a warm water bottle removed before laying the baby down. Maintain chest-to-chest contact for 60 seconds post-lay-down before beginning to withdraw. Remove the hand progressively over 2 to 3 minutes. This partial recreation of warmth and pressure reduces the startle response triggered by the temperature change of the sleep surface.

5.     Sensory bridge between arms and crib. The gap between the warmth and movement of being held and the stillness of the crib is at its widest in weeks 1 to 4. The CalmCuddle Pillow provides rhythmic movement, gentle pressure, and consistent warmth that help newborns sustain sleep through the inter-cycle transitions that most frequently cause waking at this age. Many parents introduce it in weeks 2 to 3, specifically during the period when transfer failures are most frequent. Discover how it works →

 

Frequently Asked Questions

How much should a 1-week-old sleep?

A 1-week-old typically sleeps 16 to 18 hours per 24 hours, distributed across 8 to 12 periods of 1 to 3 hours. There is no day-night pattern at this age — the circadian rhythm does not begin forming until 6 to 8 weeks. Feeding every 2 to 3 hours is biologically necessary due to the newborn's very small gastric capacity. A 1-week-old consistently sleeping more than 19 hours or fewer than 11 hours warrants a pediatric check.

 

Is it normal for a 2-week-old to sleep all day?

Yes — a 2-week-old sleeping 16 to 17 hours per 24 hours is completely normal. Day-night confusion peaks around weeks 2 to 3, meaning the longest sleep stretch may fall during the day rather than at night. This resolves gradually between 6 and 10 weeks as melatonin secretion begins establishing a nocturnal pattern. Consistent morning light exposure during the first feed of the day accelerates this process.

 

When do newborns start sleeping longer stretches?

Most newborns produce their first longer stretch of 3 to 4 hours sometime between weeks 3 and 6. By 3 months, approximately 50% of babies achieve a 5-hour uninterrupted stretch. The transition is gradual and nonlinear — expect week-to-week variation rather than steady linear improvement. The 5 to 6-week mark often brings a temporary regression before improvement resumes.

 

The Bottom Line

The chaos of newborn sleep in weeks 1 to 4 is not a problem you are causing. It is biology — a brain without a circadian clock, running on sleep cycles half the length of an adult's, spending half its sleep time in active neural construction. Every week, imperceptibly, the architecture organizes. Patterns emerge — not fast, not linearly, but they emerge.

 

The hardest part of these first four weeks is the gap between arms and crib. Discover how other parents are bridging it →

 

 

Scientific References

All sources cited are peer-reviewed publications. PMIDs verified April 2026.

 

[1] Rivkees SA (2003). Developing circadian rhythmicity in infants. Pediatrics, 112(2), 373–381. — Documents that endogenous melatonin secretion begins developing a nocturnal peak at 6 to 12 weeks postpartum — basis for the absence of day-night pattern in weeks 1 to 4.

    PubMed: https://pubmed.ncbi.nlm.nih.gov/12897290/

 

[2] Graven SN & Browne JV (2008). Sleep and brain development. Newborn and Infant Nursing Reviews, 8(4), 173–179. — Documents that 50% of neonatal sleep is active REM, and that motor activations during this phase contribute to sensorimotor development.

    DOI: https://doi.org/10.1053/j.nainr.2008.10.010

 

[3] Zangen S et al. (2001). Archives of Disease in Childhood, 85(3), 228–232. — Documents neonatal gastric capacity of ~7ml at birth expanding to 22–27ml by end of week 2 — biological basis for mandatory 2 to 3-hour feeding intervals.

    PubMed: https://pubmed.ncbi.nlm.nih.gov/11641459/

 

[4] van Sleuwen BE et al. (2007). Swaddling: a systematic review. Pediatrics, 120(4), e1097–e1106. — Documents 28% increase in uninterrupted sleep duration with swaddling; establishes rolling onset as discontinuation criterion.

    PubMed: https://pubmed.ncbi.nlm.nih.gov/17908730/

 

[5] Barr RG et al. (2009). CMAJ, 180(7), 727–733. — Population data confirming the Period of PURPLE Crying peak at 5 to 6 weeks across cultures, resolving spontaneously by 3 to 4 months. (Note: plan cited PMID 19416502 — actual publication is CMAJ PMC2659818.)

    PMC: https://pmc.ncbi.nlm.nih.gov/articles/PMC2659818/