PooPeeMlk
Get notified at launch
🌙Sleep

The 4-month sleep regression: what's actually happening

D
By a twin dad4 min readUpdated 2026-05-02

The '4-month sleep regression' isn't a regression — it's a permanent change in sleep architecture. Here's what shifts, why it disrupts sleep, and what actually helps.

Around 3–5 months, many families experience a sharp deterioration in sleep that was, until that point, getting gradually better. More frequent night waking, shorter naps, more difficulty settling — all arriving together and apparently without cause.

This period is widely called the "4-month sleep regression." The name is misleading in an important way: it's not a regression. Your baby hasn't gone backwards. Their sleep has permanently changed.

What actually changes at 4 months

Newborn sleep and adult sleep are organised very differently. Newborns fall quickly into a deep, REM-dominant sleep state and cycle through lighter and deeper phases in relatively short, loosely structured cycles of around 40–50 minutes. This is why very young babies can be moved without waking and why they often seem to sleep through everything.

From around 3–5 months, sleep architecture matures. Babies begin cycling through the same distinct stages that adults use: light NREM sleep, deeper NREM sleep, and REM sleep. These cycles become more structured — and, critically, the brief arousal between cycles that adults experience and immediately fall back through becomes something babies must learn to navigate.1

In adults, this between-cycle arousal is so brief and automatic that we don't remember it. In a 4-month-old who has just developed this new sleep structure, the arousal between cycles is a genuinely novel experience. Some babies learn to resettle into the next cycle smoothly. Others fully wake, need assistance to go back to sleep, and do this every cycle throughout the night.2

This is why the so-called regression often presents as:

  • Multiple wakings per night in a baby who was previously sleeping longer stretches
  • Shorter naps that end at the 30–45 minute mark (one cycle)
  • More difficulty settling at bedtime
  • Increased fussiness at what should be sleep time

This change is permanent

The key thing to understand is that this new sleep architecture doesn't go away. A newborn's undifferentiated deep sleep is a developmental stage that has now passed. Your baby is not going to return to sleeping the way they did at 6 weeks.

What changes is not the architecture — it's whether your baby can resettle independently between cycles. That capacity develops over weeks to months and is influenced by how the sleep environment is set up and what associations the baby has built around falling asleep.

This distinction matters because it shapes how to respond. Waiting for the "regression to pass" and expecting the previous sleep pattern to return will not work. Adjusting to the new sleep architecture will.

Typical duration of disruption

The adjustment period after sleep architecture matures typically runs 2–6 weeks for most families. This is the window during which babies are newly navigating this structure and families are adapting routines to support it.

Some families experience minimal disruption — babies who settle well between cycles from the start. Others find several months of fragmented nights before settling patterns emerge. Both are within the normal range.

What helps

Consistent wind-down routine. A predictable sequence — bath, feed, dim light, brief calm activity, put down — helps signal to the maturing sleep system that sleep is coming. It doesn't need to be long: 15–20 minutes is enough. The consistency matters more than the specific elements.

Dark room. Around 3–5 months, the circadian rhythm is consolidating. Darkness in the sleep space supports melatonin release and reinforces the day-night distinction that is still developing. Blackout blinds are useful here in ways they aren't for newborns, whose rhythms aren't yet light-responsive.

Give baby a chance to resettle. When a baby wakes between cycles, waiting 2–3 minutes before responding gives them an opportunity to settle back without intervention. This is not about sleep training — it's about not reflexively interrupting what might resolve on its own. A baby who is genuinely distressed will escalate; one who is cycling between sleep stages may not.

Drowsy but awake at bedtime. Putting a baby down when they are drowsy but not fully asleep means they fall asleep in the same environment they'll wake into between cycles. This consistency can make between-cycle resettling easier. It is harder to achieve than it sounds.

What doesn't help

Assuming every waking is hunger. Between-cycle arousals are not feeds. A baby who has recently fed, gained weight appropriately, and has good nappy output is not waking from hunger in most cases. Feeding at every waking can create a strong association between waking and feeding that extends the period of fragmented nights.

Adding new sleep associations under pressure. Bouncing, rocking, or feeding a baby to full sleep at every waking teaches the baby that this is how sleep happens. It works in the short term — the baby sleeps — but creates a dependency that must be resolved at each between-cycle arousal. It doesn't fix the underlying architecture change; it creates new management requirements on top of it.

Expecting it to "just pass." As noted above, the sleep architecture change is permanent. Waiting passively without adjusting the sleep environment or routine is unlikely to produce significant improvement.

Good to know

The 4-month change is a maturation of sleep cycles, not a temporary disruption. What varies between families is how quickly babies learn to navigate the transition between cycles independently — and that is influenced by routine, environment, and individual temperament.

For a deeper understanding of how infant sleep cycles work from birth onwards, see: Newborn sleep: why it's so weird, and what to expect

← Back to the complete guide: Safe sleep and newborn sleep

Also in this cluster: Newborn sleep cycles · Dropping naps by age

Sources

  1. NHS. "Helping your baby to sleep." NHS, 2024. https://www.nhs.uk/conditions/baby/caring-for-a-newborn/helping-your-baby-to-sleep/
  2. American Academy of Pediatrics. "Sleep-Related Infant Deaths: Updated 2022 Recommendations for Reducing Infant Deaths in the Sleep Environment." Pediatrics 150(1), 2022. https://publications.aap.org/pediatrics/article/150/1/e2022057990/188304/Sleep-Related-Infant-Deaths-Updated-2022-Recommendations
  3. NICHD Safe to Sleep. "About the Safe to Sleep Campaign." National Institute of Child Health and Human Development, 2024. https://safetosleep.nichd.nih.gov/

Footnotes

  1. American Academy of Pediatrics. "Sleep-Related Infant Deaths: Updated 2022 Recommendations for Reducing Infant Deaths in the Sleep Environment." Pediatrics 150(1), 2022. https://publications.aap.org/pediatrics/article/150/1/e2022057990/188304/Sleep-Related-Infant-Deaths-Updated-2022-Recommendations

  2. NHS. "Helping your baby to sleep." NHS, 2024. https://www.nhs.uk/conditions/baby/caring-for-a-newborn/helping-your-baby-to-sleep/

Get notified
Be first in line when we launch.
We'll email you once. No spam, no newsletter — just the launch.
Disclaimer: This is not medical advice. PooPeeMilk shares general information to help you make sense of what you're seeing. Always consult your pediatrician with concerns, especially if your baby seems unwell.
Read full disclaimer →
🌙
Next · Sleep
Back to sleep: why position matters most