Home / Sleep Guide / Sleep

The 4-month sleep regression

Why your good sleeper suddenly isn't, and the 3 small adjustments that get most babies back on track.

TL;DR The 4-month sleep regression isn't temporary. It's a permanent change to how your baby's sleep cycles are structured. They're shifting from newborn sleep (deep, dreamy) to adult-style sleep (cycles with brief surfacing). It usually starts between 3 and 5 months and lasts 2 to 6 weeks until baby learns to fall back asleep on their own. The three fixes that work: shorter wake windows, a consistent (boring) bedtime routine, and falling-asleep practice in the crib.

Want a personalized wake window schedule for your baby's exact age? Use our free wake windows calculator.

What's actually happening

Calling this a "regression" is misleading. Your baby isn't going backwards. They're going through a permanent neurological shift, and most of the "sleep regressions" later in the first year aren't really sleep regressions at all. The 4-month one is the only one that's universal and biological.

The science. Until around 3 to 4 months, babies sleep in two simple stages: light (REM-like) and deep. They cycle between them but don't fully wake between cycles.

Around 4 months, their brain matures and adds new sleep stages (non-REM 1, 2, 3), making their sleep architecture closer to an adult's. The catch: between every full sleep cycle (every 30 to 50 minutes for babies), they briefly surface into a near-waking state.

If they don't have the skill to fall back asleep, they fully wake. And cry.

Newborn baby asleep on a soft surface during the 4-month sleep transition
The 4-month regression is permanent — baby's sleep architecture is reorganizing forever. The good news: the new structure is what adult sleep looks like.

The signs

You'll notice 2 or 3 of these all at once:

  • Frequent night wakings (every 30 to 60 minutes for some babies).
  • Short naps. 30 to 45 minutes when previously they took longer.
  • Difficulty falling asleep at bedtime.
  • Early-morning wakings between 4:30 and 5:30 AM.
  • Increased daytime fussiness from cumulative sleep loss.
  • Often paired with a feeding shift (more frequent or fussier feeds).

Not sure if this is actually the 4-month regression?

Five questions tells you: 4-month regression, a later one (8/11/12/18mo, 2yr), or one of the imposters (teething, illness, schedule problem). Each result has a 4-bullet action plan.

Identify the regression →

How long it lasts

The biological change is permanent. The disruption period is usually 2 to 6 weeks, depending on what you do during it.

  • If you change nothing: 4 to 6 weeks until baby learns to self-settle from random crib practice.
  • If you adjust wake windows + add a bedtime routine: 2 to 4 weeks.
  • If you add gentle falling-asleep practice: 1 to 3 weeks.

Some babies bounce back in 7 to 10 days. Some take a full 6 weeks. The variation isn't about whether you're "doing it right." It's about your baby's individual temperament and sleep drive.

Sleeping baby photographed in soft natural light during a quiet nap
Three fixes carry the regression: full feeds during the day, drowsy-but-awake at bedtime, and consistent wake windows of 1.5-2 hours.

The 3 fixes that work

Fix 1: Shorten wake windows to 1.5–2 hours

The single most common mistake at the 4-month regression is keeping wake windows at the 1.5–2.5 hour range that was working before. Babies' sleep needs shift around now. Most 4-month-olds need wake windows of:

  • Morning wake to first nap: 1.5 hours
  • Between subsequent naps: 1.75–2 hours
  • Last nap to bedtime: 2–2.25 hours (slightly longer)

If your baby is wired and crying at bedtime, they're almost certainly overtired. Shorten the last wake window by 15 minutes. Counterintuitive, but it's the fix.

Get a personalized wake window schedule

Enter your baby's age and morning wake time. Get a sample schedule with nap times and bedtime in 30 seconds.

Try the calculator

Fix 2: A boring, predictable bedtime routine

Babies at 4 months can recognize routine. They use it to know sleep is coming, which preps the brain to release sleep-onset hormones.

The routine doesn't need to be elaborate. 10 minutes is fine. The point is doing it the same way every night:

  • Bath (every other night, or every night if it doesn't make baby wired).
  • Pajamas and sleep sack on.
  • Dim the lights.
  • Turn on white noise.
  • Final feed in dim light, not in the crib.
  • One short song or lullaby.
  • Crib.

Boring is the goal. Don't make bedtime a party.

Fix 3: Falling-asleep practice in the crib

This is the one that actually shortens the regression. Your baby needs to learn how to fall asleep without being held, rocked, or fed to sleep. Otherwise every brief night-waking requires you to recreate the conditions they fell asleep in.

You don't need to do formal sleep training (cry-it-out, Ferber, etc.), though you can if you want. The simpler approach:

  1. At bedtime, do the routine, feed, then put baby in the crib drowsy but awake.
  2. If they cry, pick them up after 2 to 3 minutes, calm them, put them back down still awake.
  3. Repeat. They'll fall asleep eventually. Each night this gets faster.

This isn't sleep training. It's giving baby the chance to practice. Some babies need 1 to 2 nights. Some need 1 to 2 weeks. If you're not seeing improvement after a week of consistent practice, formal sleep training methods become the next option.

What doesn't work (don't do these)

  • Adding a dream feed. Usually doesn't fix night wakings caused by the regression. Sometimes makes it worse.
  • Letting baby sleep on you. Patches the symptom, doesn't address the cause. Hard to undo later.
  • Co-sleeping as a fix. Fine if it's a permanent choice. If you start during the regression "just to survive," you've signed up for a different long-term arrangement.
  • Loud rocking, walking, or driving to sleep. Same issue. Patch, not fix.
  • Sleep training before 16 weeks. Sleep researchers generally recommend waiting until at least 16 weeks (4 months adjusted age) for any cry-it-out method.
Baby curled up sleeping peacefully on a neutral blanket
Daytime feeds matter more during the regression than ever. A baby fed every 3 hours during the day doesn't need a 2 AM snack.

What about feeding?

The 4-month regression often comes with a feeding shift. Babies may want to feed more often because they're using feeds to fall back asleep. This isn't always real hunger.

The test: if baby takes a full feed, it's hunger. If baby takes 30 seconds of suck and conks out, they're using the breast or bottle as a sleep tool.

If it's the latter, gradually shift toward falling-asleep-without-feeding. Daytime first, then bedtime, then night wakings. See paced bottle feeding for the bottle technique that helps separate feeding from sleeping.

When to call your pediatrician

  • Sleep disruption lasts more than 6 weeks despite consistent adjustments.
  • Baby is fussy or crying excessively during the day, not just at sleep.
  • Weight gain has slowed.
  • You suspect ear infection (extra fussy lying down) or reflux (arching back during feeds).
  • You're not sleeping enough yourself and feeling severely depressed or anxious. Postpartum mental health is a real factor here. Tell your provider.

Sources

Questions parents ask

Pulled from Google's "People Also Ask" box for this topic, answered by our editors with the research and our test-family notes.

How long does the 4-month sleep regression last?

The disrupted nights typically last 2 to 6 weeks. The underlying shift — your baby moving from newborn sleep cycles into adult-style cycles with light-sleep wake points — is permanent. So the chaos is temporary, but the new sleep pattern is forever.

Is there really a 4-month sleep regression?

Yes. It is a normal neurological shift that happens between roughly 12 and 18 weeks. Sleep architecture matures, brain waves change, and babies start surfacing through light-sleep cycles every 45 to 90 minutes. Not every baby has dramatic disruption, but the change happens to all of them.

What is the 5-3-3 rule for 4-month-olds?

A schedule shorthand: about 5 hours of total awake time across the day, 3 hours between the last nap and bedtime, and 3 nighttime wakings considered normal before you reset. Useful as a check, not a prescription. Some 4-month-olds do better on shorter wake windows.

How do I know if my 4-month-old is having a sleep regression?

Sudden short naps under 45 minutes, frequent night wakings every 1 to 2 hours, fights at bedtime, fussy mornings — all happening at once, without illness, teething, or a major schedule change. Our Sleep Regression Identifier quiz walks through it in 60 seconds.

Keep reading

Free Tool
Is this actually a regression? (Quick quiz)
Free Tool
Wake windows calculator (by age)
Sleep · Reference
Wake Windows by Age (Free Printable)
Sleep · Next
The 8-month sleep regression
Sleep · How-to
How to Drop the Swaddle Safely
Sleep · Compare
Sleep training methods compared
Feeding · Explainer
Cluster Feeding Decoded