4-Month Sleep Regression: What It Is and How to Survive It

You finally felt like you were getting the hang of it. Your baby was sleeping in 3–4 hour stretches, maybe even a longer chunk at night. You'd stopped dreading bedtime quite as much. You might have even dared to hope that sleep was getting better.

Then somewhere around 3.5 to 4 months, it stopped. Your baby is up every 45 minutes, won't settle without being held, and you're back in the thick of survival mode — wondering what happened and whether it will ever end.

You're not imagining it. You didn't do anything wrong. This is the 4-month sleep regression — one of the most disruptive and misunderstood events in the first year of parenting. Here's what you need to know.

---

What Is the 4-Month Sleep Regression?

The 4-month sleep regression is a permanent change in how your baby sleeps — not a temporary disruption that passes.

Around 3.5–4 months, your baby's sleep architecture undergoes a major neurological shift. Up until now, your baby has been cycling through just two sleep stages: active sleep (similar to REM) and quiet sleep (similar to deep, non-REM sleep). These cycles are relatively short and shallow.

At 4 months, the brain matures and sleep reorganizes into a more adult pattern — with four distinct stages:

• Stage 1: Light sleep / drowsiness
• Stage 2: Stable light sleep
• Stage 3: Deep sleep (hardest to rouse from)
• Stage 4: REM sleep / active dreaming

This is developmentally excellent news. It means your baby's brain is maturing on track.

Here's the catch: with more stages comes more transitions between stages — and at each transition, your baby rouses slightly. Adults do this too, but we've learned to roll over, adjust, and drift back to sleep without fully waking. Your baby hasn't learned this yet. Every time they cycle through to a light sleep stage, they come to enough awareness to realize they're not where they were when they fell asleep — not in your arms, not at the breast, not being rocked. And they cry.

This is why the 4-month regression is different from every other regression. Most sleep regressions are caused by temporary developmental changes — a growth spurt, a new skill, a developmental leap. Those pass, and sleep returns to the prior baseline. The 4-month regression doesn't work that way. The new sleep architecture is permanent. Your baby's sleep will not revert to what it was before.

That sounds alarming, but here's what it actually means: the regression isn't a problem to be endured until it's over. It's a signal that your baby's sleep has changed, and your approach needs to adapt with it.

---

Signs the 4-Month Regression Is Happening

You'll know it when it hits, but here's what to watch for:

Sleep changes:

• Waking every 45–90 minutes at night after previously sleeping longer stretches
• Short naps (30–45 minutes) when they used to nap longer
• Difficulty falling asleep at bedtime despite obvious tiredness
• Needing more help to fall asleep than before (more rocking, more nursing, more contact)

Daytime changes:

• Increased fussiness and irritability
• Seeming more alert and interested in the environment than before
• Appetite changes — sometimes more, sometimes erratic
• Developmental surges: you may notice your baby starting to roll, bat at objects, or become more engaged and social

Timing: The regression typically kicks in between 3.5 and 4.5 months. Some babies hit it earlier (3 months), some a bit later (5 months). If your baby was premature, count from their corrected age.

---

Why the 4-Month Regression Hits Hardest

Of all the sleep regressions in the first two years, parents consistently report the 4-month regression as the most brutal. Why?

1. You're at your most exhausted. Three to four months in, the accumulated sleep debt of the newborn phase is real. You were just starting to feel human again.

2. It lasts the longest if nothing changes. Because the cause is permanent (new sleep architecture), many families find themselves in extended disrupted sleep — weeks or months — if they don't adjust their approach. A growth spurt regression resolves in 3–7 days whether you do anything or not. The 4-month regression doesn't.

3. Sleep associations suddenly matter. Before 4 months, your baby could be nursed, rocked, or bounced to sleep and you might not notice problems. Now, with more frequent arousals, your baby needs whatever they used to fall asleep initially to get back to sleep at every wake. If that was you — your presence, your breast, motion — that's now happening 8–10 times a night.

---

How Long Does the 4-Month Sleep Regression Last?

Here's the honest answer: it depends entirely on what you do about it.

If your baby already knows how to fall asleep independently (without being nursed, rocked, or held to sleep), the regression may last only 2–4 weeks as they adjust to the new sleep architecture.

If your baby relies on a sleep association to fall asleep — which is most babies at 4 months — the disruption can continue indefinitely, because the underlying issue (needing parental help to fall back to sleep at every arousal) isn't going away on its own.

This is why 4 months is often considered the ideal time to begin sleep training, if your family is ready. The developmental changes that caused the regression are now complete and stable, so you're working with your baby's new sleep biology rather than fighting against it. For a full breakdown of approaches, see our guide on baby sleep training.

---

5 Survival Strategies for the 4-Month Sleep Regression

Whether you're ready to sleep train or not, these strategies will help you get through the hardest weeks.

Strategy 1: Nail the Wake Windows

At 4 months, wake windows are typically 1.5–2 hours. A baby who's overtired at bedtime will fight sleep harder and wake more frequently. A baby put down too early won't have enough sleep pressure to stay asleep.

For age-appropriate wake windows and nap schedules, see our baby nap schedule guide — timing is one of the most powerful (and underused) sleep tools available.

Strategy 2: Prioritize the Sleep Environment

Before trying anything else, make sure the basics are in place:

• Darkness: The room should be as dark as possible. Even small amounts of light can disrupt sleep at this stage of circadian rhythm development.
• White noise: A consistent white noise machine running through the night helps mask environmental sounds that cause arousals. Aim for 65–70 dB — slightly louder than a quiet conversation.
• Appropriate temperature: 68–72°F is the ideal sleep temperature for most babies. Overheating is both a safety concern and a sleep disruptor.

Strategy 3: Practice "Drowsy but Awake"

Start putting your baby down in their sleep space when they're drowsy but still awake — not fully asleep. This is the foundational skill of independent sleep. It may not work at first. That's okay. The practice matters.

Even if your baby falls asleep in 1 out of 10 drowsy-but-awake attempts, that's one less wake per night where they need full resettling. Over time, the ratio improves.

Strategy 4: Give the Dream Feed a Try

A dream feed — a brief, sleepy feed around 10–11 PM before you go to bed — can extend your baby's longest sleep stretch into the early morning hours, giving you a meaningful chunk of sleep before the 4 AM wake window hits. Not every baby responds, but for many families it's a game-changer in the regression weeks.

Strategy 5: Take Turns

Survival mode is a team sport. If you have a partner, divide the night into defined shifts rather than both responding to every wake. A clear handoff — you take 11 PM–3 AM, I take 3 AM–7 AM — means at least one person is getting a solid stretch of sleep. That person functions better tomorrow, and tomorrow night, you switch.

Single parents: consider whether there's a family member or trusted friend who could take a night shift once or twice a week during the worst of it. Asking for help is not failing.

---

What NOT to Do During the 4-Month Regression

Don't try every new approach every night. Inconsistency is harder on your baby than any single imperfect approach. If you're nursing to sleep, nurse consistently. If you're trying to do some amount of drowsy-but-awake, do it consistently. Switching methods every night teaches your baby nothing — it just extends the confusion.

Don't panic-buy every sleep gadget. Motion swings, vibrating bassinets, and sleep soothers may help at the margins, but they don't fix the underlying issue. If your baby is waking every 45 minutes and the answer is always motion or nursing, you haven't taught them anything about falling back to sleep independently.

Don't compare to other babies. Your neighbor's baby might have sailed through 4 months. Your baby might be waking 12 times a night. Both are within the range of normal. Comparison makes this phase harder than it needs to be.

---

When Does the 4-Month Sleep Regression End?

For babies who develop independent sleep skills — either naturally or through sleep training — the regression typically resolves within 2–4 weeks of beginning the new approach. After a week of sleep training (if you choose to go that route), most families report dramatically improved sleep within 5–7 nights.

For families who don't address sleep associations, the disruption can continue for months. Many parents of 8- and 10-month-olds who are still waking frequently trace the problem back to the 4-month regression and the habits that solidified during it.

The good news: it's never too late. Sleep training works at 4 months, 6 months, 8 months, and beyond. Earlier is generally easier, but later is still very much possible.

---

A Note to Exhausted Parents

If you've read this far at 2 AM, holding a wide-awake 4-month-old while you're running on four hours of broken sleep, I want to say something directly: you are doing an extraordinarily hard thing.

The 4-month regression is genuinely one of the hardest stretches of new parenthood. The sleep deprivation is real, the stakes feel high, and the solutions require energy and consistency that are hard to summon when you're this tired.

But this phase does pass. Most families report that once they find an approach that works and commit to it, things improve within 2–3 weeks. You are closer to the other side than it feels right now.

---

---

FAQ: 4-Month Sleep Regression

Should I sleep train during the 4-month regression?

Yes, if you're ready — and in fact, 4 months is one of the best times to start. The neurological changes that caused the regression are now stable, and you're working with your baby's new sleep architecture rather than fighting against it. Waiting doesn't make sleep training easier.

My baby is 5 months and still waking constantly. Is this still the regression?

Possibly. If your baby developed sleep associations during the regression (needing to be nursed or rocked to sleep), those associations are keeping the waking going even after the regression itself has passed. Addressing the sleep associations — through sleep training or gradual fading — will resolve the waking.

My baby won't nap. Is that part of the 4-month regression?

Yes. Short naps (30–45 minutes) are extremely common during this period for the same reason as night waking — your baby arouses between sleep cycles and can't independently fall back to sleep. Work on drowsy-but-awake placement and consistent wake windows, and nap length often improves alongside nighttime sleep.

Will things go back to normal once the regression is over?

Not to the old normal — the new sleep architecture is permanent. But once your baby learns independent sleep skills, they will sleep well in the new architecture. Many families find that post-regression sleep is actually better than pre-regression sleep, because babies are now capable of more consolidated nighttime sleep if taught the skills to do so.