4 Month Sleep Regression: Why It Happens, How Long It Lasts, What Actually Helps

If you’re reading this at 2am because the baby who used to sleep stretches is suddenly up every 45 minutes, you’re not imagining it and you didn’t break anything. The 4 month sleep regression is real, it’s biological, and across more than 68,000 families in our research (68,366 to be exact), it’s one of the most predictable disruptions of the first year. The good news: it’s also one of the most temporary.

Before we go further, the question you came here to answer:

What is the 4 month sleep regression?

The 4 month sleep regression is the period – usually somewhere between 3 and 5 months – when a baby’s sleep stops looking like newborn sleep and starts looking like adult sleep. Newborns have two broad sleep states. By around 3 to 4 months, those consolidate into the four-stage cycle the rest of us use for the rest of our lives, with lighter stages, deeper stages, and full arousals between cycles.¹

“Regression” is the wrong word, honestly. Nothing is going backward. Your baby is doing exactly what their brain is supposed to do at this age. The reason it feels like regression is that the new cycles end every 45 to 60 minutes, and your baby now has to figure out how to bridge those arousals back into sleep. Until they do, you’re the bridge.

Why it happens

Three things are shifting at once around 3 to 4 months, and they all hit sleep:

• Sleep architecture matures. The biphasic newborn pattern (active sleep / quiet sleep) reorganizes into the four-stage NREM/REM cycle. Your baby now spends more time in lighter sleep stages, which means more opportunities to wake fully.¹
• Circadian rhythm consolidates. Melatonin production and the day/night distinction are still settling. Most babies don’t have a fully adult-like circadian pattern until somewhere between 3 and 6 months.
• Cognitive load increases. The same month, babies are doing a lot of new work – social smiles, head control, hand-eye coordination, sometimes early rolling. The brain processes that work overnight, which can mean lighter, more fragmented sleep.

None of these are problems. They’re developmental milestones in the sleep system. That’s also why the change underneath is permanent – sleep cycles don’t un-mature.

Why your baby suddenly wakes at the end of every cycle

Signs your baby is in the 4 month sleep regression

In our family research, the signs that line up most consistently are:

• Night wakings increase, often sharply, in a baby who was previously doing longer stretches.
• Naps shorten to roughly one sleep cycle (about 30 to 45 minutes) and stay there.
• Bedtime gets harder – the baby who used to drift off now fights it.
• Early morning wakings start, often between 4 and 5am.
• Fussiness ramps up around every sleep transition, not just the hard ones.
• Day/night patterns wobble briefly even if your baby had them locked in before.

You don’t need all of them. Two or three, showing up at the same time around the right age, is usually enough.

What our data shows

Across more than 68,000 families in our 2026 dataset, a few patterns repeat enough to be worth naming:

• The vast majority of families saw nights stabilize within 2 to 6 weeks once they stopped layering new sleep interventions on top of each other. The single biggest predictor of a longer regression in our data wasn’t the baby – it was how many things the parents changed in week one.
• Families who moved bedtime 15 to 30 minutes earlier and protected an age-appropriate wake window reported faster resolution than families who held their pre-regression schedule.
• Most families who described the regression as “never-ending” were inside the typical 2 to 6 week window when they said it. Sleep deprivation distorts time.
• Families who tried to start formal sleep training in the middle of the regression were more likely to report it “didn’t work” than families who waited until the regression stabilized and then trained. Same methods, different outcomes by timing.

“Across 68,366 families in our research, the most predictable thing about the 4 month regression isn’t how bad it gets. It’s how often parents try the wrong fix first. Wake windows and an earlier bedtime do more in week one than any program.”

If you’re Googling “is this normal” at 2am, the honest answer from the data is: almost always yes. The internet has not been helping. AI-generated sleep advice often gets the basics wrong, and the “your baby’s sleep is broken” framing in most sleep app marketing this season is not what the data supports.

If you’d rather see where your specific family sits relative to that dataset than read more general guidance, that’s what the Betteroo quiz was built for.

What actually helps

Five moves do more, in our data, than any single sleep program during this window. None of them are heroic. All of them are boring on purpose.

1. Protect the wake windows.

At 4 months, most babies handle about 60 to 90 minutes of awake time between sleeps. Push past that and you get an overtired baby, who sleeps worse, not better. If you’re unsure where your child sits, the baby sleep schedule by age guide breaks down the typical windows month by month.

Typical awake-time tolerance between sleeps. The 4-month row is the one to memorize this week.

Age	Wake window	Typical naps	Total day sleep
3 months	45 to 75 min	4 to 5	14 to 17 hours
4 months  ⭐	60 to 90 min	3 to 4	12 to 16 hours
5 months	75 to 105 min	3	12 to 15 hours
6 months	90 to 120 min	2 to 3	12 to 15 hours
7 months	2 to 2.5 hours	2 to 3	12 to 14 hours

Ranges are typical; healthy babies vary. If your baby consistently sits at the high or low end, that’s usually fine. See the full wake windows by age guide for nuance.

2. Move bedtime earlier.

Counterintuitively, an earlier bedtime usually means fewer night wakings, not more. Overtired babies wake more often and earlier in the morning. Try moving bedtime 15 to 30 minutes earlier than your pre-regression baseline and hold it there for at least four nights before judging.

3. Pick one consistent fall-asleep cue and keep it.

Rocking, feeding, pacifier, contact, sound machine – any one of these can work. The problem isn’t which cue you pick. It’s switching between three different cues across a single night because everything feels broken. Pick one. Use it for 7 to 10 nights. Then evaluate.

4. Don’t introduce new sleep crutches you’re not willing to keep.

If you start nursing your baby fully to sleep in week one of the regression because nothing else works, that’s a reasonable survival move – as long as you’re prepared to keep doing it for a while. Adding a new sleep crutch under desperation and then trying to remove it three weeks later is one of the most predictable sources of a second, harder disruption.

5. One adult sleeps first.

If there are two adults in the house, take shifts. One person goes to bed at 8:30pm and handles 8:30 to 2am. The other handles 2am to morning. The version of you who has slept four straight hours is a more competent parent than the version of you who is co-rocking and co-anxious from 11pm to 4am. This isn’t optimization – it’s recovery. Parent sleep loss measurably impacts maternal mood and parenting stress, and the loss compounds across weeks if one adult absorbs all of it.⁶

What doesn’t help (and what to skip this week)

Equal time on the things that aren’t going to move the needle and might make it worse:

• Schedules from social media. The viral “perfect 4 month schedule” you saved at 3am was almost certainly written by someone who doesn’t know your baby. Your wake windows and your baby’s temperament matter more than any shareable PDF.
• Adding rice cereal to bottles. Not a sleep fix. Pediatric guidance has long advised against this as a sleep intervention – it’s a choking-and-aspiration risk, and the evidence that solids before 6 months change night wakings is weak at best.
• Changing five things at once. If you change the room temperature, the swaddle, the bedtime, the feed schedule, and the fall-asleep cue in the same week, you won’t know which one helped or hurt. Pick one variable.
• Starting full sleep training mid-regression. The major evidence base for structured behavioral sleep training – Mindell’s 2006 review and Hiscock’s 2007 trial – looked at infants 6 months and older.³ ⁴ Four months is biologically early for most methods. Wait two to four weeks for the regression to stabilize, then decide.

The 5/5 Rule (Betteroo’s named framework)

If you remember nothing else from this guide, remember this: five moves help, five myths don’t. We call it The 5/5 Rule. It’s the shortest, hardest-tested distillation of what 68,000+ families taught us about the 4 month regression. Screenshot it for the 3am moment.

Five moves that actually help, and five myths that quietly waste your week.

DO THIS 5 moves that help

1. Protect 60 to 90 min wake windows. Overtired = worse sleep, not better.
2. Move bedtime 15 to 30 min earlier and hold it for 4 nights.
3. Pick one fall-asleep cue. Use it for 7 to 10 nights before evaluating.
4. Don’t add a new sleep crutch you’re not willing to keep for weeks.
5. Adults take shifts. One sleeps 8:30 to 2am, the other 2am to dawn.

SKIP THIS 5 myths to ignore

1. Viral “perfect” schedules. Written by someone who doesn’t know your baby.
2. Rice cereal in bottles. Not a sleep fix. Choking and aspiration risk.
3. Changing 5 things in one week. You won’t know what helped or hurt.
4. Starting full sleep training mid-regression. Major studies looked at 6+ months.
5. Assuming it’ll last forever. Most families stabilize within 2 to 6 weeks.

Screenshot this for the 3am moment when you need a quick gut check.

When it’s NOT just the regression

The 4 month sleep regression is common, but it’s not the only thing that can disrupt sleep at this age. Talk to your pediatrician – not your sleep app – if any of the following are happening:

• Persistent breathing changes during sleep – snoring with pauses, unusually heavy breathing, mouth breathing that won’t resolve. AAP safe sleep guidance directly addresses this as a flag worth raising with a pediatrician.²
• Total daily sleep dramatically below the AASM range (roughly 12 to 16 hours including naps at this age) across multiple weeks, not single bad nights.⁵
• Missed developmental milestones alongside the sleep change.
• Signs of reflux or feeding pain that escalate at bedtime – arching, crying that doesn’t soothe in horizontal positions, frequent spitting up with discomfort.
• A sudden change in alertness, tone, or temperature.

Outside of those, what you’re experiencing is almost certainly the regression. Tiring, but typical.

A note on sleep training during the regression

The desperation to start sleep training mid-regression is real and we don’t blame anyone for feeling it. But the evidence base doesn’t support 4 months as the right window for most families. Mindell’s landmark 2006 review and Hiscock’s long-running follow-up trials both looked at babies 6 months and older.³ ⁴ That doesn’t mean nothing helps at 4 months – the five moves above all help. It means the formal “methods” you’ve heard of were designed and tested later.

If you do want to compare approaches once your baby is closer to 6 months, the common sleep training methods guide walks through what the research actually says about each one.

Frequently asked questions

Key takeaway

The 4 month sleep regression is the most predictable, most temporary, most misnamed disruption of the first year. It’s not a defect, you didn’t cause it, and you can’t fix it – because it’s not broken. Protect the wake windows, move bedtime earlier, pick one cue, don’t stack new crutches, and take shifts. In our 68,000-family data, that’s most of what separates a 2 week regression from a 6 week one.