The 4-month sleep regression: why it happens

Why does my baby's sleep suddenly change at 4 months?

Around 4 months, according to Cleveland Clinic, your baby's sleep architecture (sleep pattern) goes through a real, lasting shift. Newborn sleep is mostly deep sleep. But around this age, babies start cycling through alternating phases of deep and light sleep. This pattern is much closer to how adults sleep, moving between lighter and deeper stages throughout the night. Cleveland Clinic identifies this shift as the mechanical cause behind the regression.

That change matters because light sleep brings more natural stirring and partial waking between cycles. An adult sleeper rolls over, resettles, and drifts back off without ever fully waking up. A baby who has only known deep sleep hasn't practiced that skill yet. So those transitions between cycles can turn into full wake-ups. This is especially true if your baby is used to falling asleep a particular way, like being nursed or rocked all the way down. The regression isn't a step backward. It's a permanent upgrade to a more mature sleep pattern that a young baby doesn't yet know how to handle on their own.

Why do wake windows matter so much now?

Cleveland Clinic defines a wake window as the amount of time your baby is awake between naps. This number shifts quickly in the early months. A newborn's wake window is roughly 30 minutes to an hour. But by 3 to 4 months, babies can typically handle about 1.25 to 2.5 hours of awake time before they're ready to sleep again.

That's a significant jump. It's easy to miss if you're still working off a newborn-era rhythm. Putting a 4-month-old down for a nap too soon, or letting them stay up too long relative to this new window, can make an already lighter sleeper even harder to settle. Watch the clock alongside your baby's own cues, rather than defaulting to the old newborn schedule. It's often one of the simplest adjustments during this stretch.

Is this regression permanent?

It's worth repeating: Cleveland Clinic characterizes the 4-month regression as typically temporary, usually lasting about one to two weeks before sleep gets back on track. For most families, the roughest stretch is genuinely short. It doesn't always feel that way at 3 a.m. on night nine, though.

What sometimes gets mistaken for 'permanent' is that a baby's sleep associations can shift during the regression and linger afterward. Say a baby who previously fell asleep independently starts needing to be nursed, rocked, or held to sleep through this stretch. That new habit doesn't automatically disappear once the regression itself resolves. That's often the real reason a two-week regression can feel like it never ends. The sleep pattern has already settled into its new form, but the newer bedtime habits haven't.

How do I help my baby learn to fall asleep on their own?

One of the most consistently recommended strategies for this stage comes from AAP-affiliated guidance: put your baby down drowsy but not fully asleep, rather than waiting until they're already asleep in your arms. A baby who learns to complete that last step — drifting off in the crib rather than being carried across the finish line — is generally better equipped to resettle on their own. This helps during the lighter-sleep stirrings that come with the new sleep pattern.

This doesn't have to be an all-or-nothing shift overnight. Many families start by simply putting their baby down a little earlier in the drowsiness process at bedtime. They build from there.

What does the research say about sleep training?

For parents considering a more structured approach, a peer-reviewed review in American Family Physician describes graduated extinction. This is a specific protocol where you put your baby to bed while still awake. Then you wait progressively longer intervals — for example, 2, then 4, then 6 minutes — before checking in.

The same review found that graduated extinction and sleep fading both reduced how long it took children to fall asleep. They also reduced how often children woke overnight, compared with a control group. Just as important for anxious parents: the review found no difference in child-parent attachment security among the groups studied. This means there was no evidence that these approaches harmed the parent-child bond. Whether and when to try any sleep-training approach is a personal decision. But it doesn't have to be made on the assumption that it will cost you closeness with your baby — the evidence doesn't support that.

What are the safe sleep basics, and when should I call the pediatrician?

Whatever approach you take to the regression itself, safe sleep guidelines don't change. NICHD's Safe to Sleep guidance is clear: place your baby on their back for every sleep, on a separate firm, flat, level surface that's free of soft objects, pillows, and blankets. This is the best way to reduce the risk of SIDS (sudden infant death syndrome). NICHD also recommends room-sharing, with your baby sleeping in your room but on their own separate sleep surface, for at least the first 6 months. This is considered safer than either bed-sharing or having your baby sleep alone in a separate room.

Call your pediatrician if the disrupted sleep doesn't seem to be resolving, especially past the typical one-to-two-week window. Also call if something about your baby's sleep, feeding, or overall behavior just feels off to you. Cleveland Clinic specifically advises checking in with your pediatrician in these situations, since it's worth ruling out other causes rather than assuming every rough stretch of sleep is simply the 4-month regression. Trust that instinct. You know your baby, and a quick call costs nothing.

Frequently asked questions

Sources

1. The AAP-affiliated guidance recommends putting babies down drowsy but not fully asleep, rather than waiting until they are already asleep, so they learn to fall asleep independently. — HealthyChildren.org (AAP)
2. Around 4 months, babies' sleep architecture shifts from mostly deep sleep toward cycling through alternating phases of deep and light sleep, similar to the adult sleep pattern — this is the mechanical driver of the regression. — Cleveland Clinic
3. Cleveland Clinic characterizes the 4-month sleep regression as typically temporary, lasting about one to two weeks before sleep gets back on track. — Cleveland Clinic
4. Cleveland Clinic advises parents to consult their pediatrician if the sleep regression doesn't seem to resolve or something feels off, to rule out other causes before attributing disrupted sleep solely to the regression. — Cleveland Clinic
5. Cleveland Clinic defines wake windows as the amount of time babies are awake between naps, with the expected wake window for babies aged 3-4 months being roughly 1.25 to 2.5 hours (compared to just 0.5 to 1 hour for newborns). — Cleveland Clinic
6. NICHD's Safe to Sleep guidance states that placing a baby on their back for every sleep, on a separate firm, flat, level sleep surface free of soft objects, pillows, and blankets, is the best way to reduce SIDS risk. — NICHD / Safe to Sleep
7. NICHD recommends room-sharing (baby sleeping in the parents' room but on a separate surface) for at least the first 6 months, noting it is safer than either bed-sharing or having the baby sleep alone in a separate room. — NICHD / Safe to Sleep
8. A peer-reviewed review in American Family Physician describes graduated extinction as a specific protocol where the parent puts the child to bed still awake and waits progressively longer intervals (e.g., 2, then 4, then 6 minutes) before checking on the child. — American Academy of Family Physicians (American Family Physician)
9. That same review found graduated extinction and sleep fading both reduced time to fall asleep and number of night awakenings compared to a control group, and did not find any difference in child-parent attachment security among treatment groups, indicating no evidence of harm to attachment. — American Academy of Family Physicians (American Family Physician)