4-Month Sleep Regression: How Long Does It Last?

Key takeaways

• Most babies work through the 4-month sleep regression in 2 to 6 weeks, though some settle in under two and others take closer to eight.
• The 4-month regression is permanent: newborn sleep architecture is replaced by adult-like sleep cycles, so the old pattern never returns.
• End-signals include night wakings dropping to 1-2 per night, one 4+ hour sleep stretch, and naps lengthening past 30-45 minutes.
• Still waking every 60-90 minutes six weeks in, or wakings paired with feeding refusal or poor weight gain, warrants a pediatrician conversation.
• Putting the baby down drowsy but awake at bedtime and keeping routines boring prevents desperation habits from outlasting the regression.

Most babies work through the 4-month sleep regression in 2 to 6 weeks, though some settle in under two and others take closer to eight. Here's the part most articles skip: the old newborn sleep pattern never comes back. Your baby's sleep cycles are maturing permanently, and the end-signals checklist below shows you when the new, more adult-like pattern has stabilized.

How long does the 4-month sleep regression last?

Most babies take 2 to 6 weeks to settle after the 4-month sleep regression begins, though published ranges vary because sleep researchers define the onset and end differently and few studies track it directly. Treat 2 to 6 weeks as the typical window, not a promise.

The typical 2-6 week window

For most babies, the roughest stretch is the first 2 weeks, when night wakings multiply and naps shorten. Things usually stabilize within 6 weeks as your baby's sleep cycles mature into adult-like patterns of light and deep sleep. This is not a phase that "ends" so much as a permanent change in how your baby sleeps.

Why your baby's timeline may differ

Your baby's timeline depends on temperament, feeding method, and whether sleep habits formed before the regression still work afterward. A baby who needed heavy rocking to sleep before week 14 often takes longer to settle than one who could fall asleep drowsy but awake. The American Academy of Pediatrics describes the 3-5 month period as a time of significant sleep-cycle maturation (HealthyChildren.org), which is why some babies breeze through in days while others take the full 6 weeks.

If you log night wakings and sleep stretches, you'll see the trend line yourself. A simple tracker like NoobieNest: Calm Baby Tracker makes that a one-tap job at 3 a.m., so you can judge whether things are improving week over week.

Why does the 4-month sleep regression happen at all?

Newborn sleep vs mature sleep cycles

Newborns sleep in two states only: active sleep and quiet sleep. They drift straight into active sleep, so they can fall asleep easily at the breast or bottle, but their sleep is shallow and breaks in short stretches. Around 3 to 4 months, the brain rebuilds sleep into the adult-style structure: distinct cycles with stages of light sleep, deep sleep, and REM, each cycle lasting roughly 40 to 60 minutes in infants.

That restructuring is the regression. Your baby now surfaces briefly at the end of every cycle instead of staying in one long stretch of shallow sleep. A baby who previously slept a four-hour block may start waking every 45 to 60 minutes, because every cycle transition is now a moment where they can wake, and they often need the same conditions they fell asleep with to fall back. It looks like a setback, but the wakings are a byproduct of a brain upgrading its hardware. The timing varies by a few weeks in either direction, and the change can arrive suddenly, which is why it feels like sleep collapsed overnight.

Why the change is permanent, not temporary

The word "regression" is misleading. Your baby is not going backward to an earlier stage and then returning; the newborn sleep architecture is gone for good. What resolves over two to six weeks is not the new cycle structure but the skill gap around it: your baby learning to link cycles together and fall back asleep without full help. That is why the answer to "how long does it last" is a range, not a date, and why the wakings that remain after the rough patch are a feature of mature infant sleep, not a phase that failed to end.

What signs show the 4-month sleep regression is ending?

The clearest signal is time: the disruption typically runs 2 to 6 weeks, so if night wakings started around week 17, a return to longer sleep stretches by week 20 to 23 is on schedule (Sleep Foundation's overview of infant sleep). But duration alone is weak evidence. The end-signals below, checked against your own log, are stronger.

End-signals checklist

• Night wakings drop back to 1-2 per night, and your baby resettles within 10-15 minutes instead of full wakefulness.
• One sleep stretch reaches 4+ hours again, even if it is not the same stretch every night.
• Naps lengthen past the 30-45 minute cycle, meaning your baby can link sleep cycles during the day, not just at night.
• Falling asleep at bedtime takes under 20 minutes with your usual routine.
• You stop seeing 4-5 a.m. "party" wake-ups where the baby is fully alert and chatty.

How to use a sleep log to spot the shift

Trends matter more than single nights, and a log is the only honest way to see them. Compare week over week: total night sleep, longest stretch, and number of wakings. Two consecutive weeks trending the right direction beats one good night.

If you are logging in Calm Baby Tracker, the sleep history screen shows exactly these three numbers, and the doctor PDF exports them for a well-child visit. Honest limitation: it will not diagnose anything or tell you when the regression ends. It just makes the week-over-week comparison visible instead of a 3 a.m. guess.

When does the regression become a sleep problem instead of a phase?

The 4-month regression is a permanent change in how your baby's sleep cycles work, not a phase that ends. So the real question is whether night wakings stay frequent well past the adjustment window, and whether nothing you try moves the needle.

Red flags past the 6-week mark

If your baby was waking 2-3 times a night and is still waking every 60-90 minutes six weeks after the bump started, that pattern deserves attention rather than patience. The same goes for wakings paired with feeding refusal, poor weight gain, or sleep stretches that never exceed 45 minutes day or night. Those signs suggest something mechanical, like reflux or a scheduling mismatch, not normal cycle maturation.

Habits formed during the regression that stick

Desperation habits are the bigger risk. Rocking, bouncing, or nursing to sleep for six straight weeks can become the only way your baby falls back asleep, and that association outlasts the regression itself. If you now spend 40 minutes resettling after every wake, the phase ended but the habit stayed. That is the point where a consistent approach to independent falling asleep is worth addressing deliberately.

What can parents do during the regression without creating new sleep habits?

Protect the first stretch of the night

The first stretch is the one most worth defending. Put the baby down at the start of the night when drowsy but still awake, so falling asleep alone stays a practiced skill rather than something that must be relearned after the regression. If you feed or rock to sleep at bedtime, you will likely need to repeat that at every 2 a.m. wake-up for months, not weeks.

During night wakes, keep it boring: dim light, minimal talk, a feed if one is due, back down. Respond consistently, but respond with the least intervention that works.

Keep routines boring and consistent

A short, identical wind-down each night is low-regret: it costs nothing and survives the regression. Resist adding new props now, because habits formed during a rough fortnight are hard to unwind. The same discipline matters later; regressions around 15, 18, and 20 months arrive with developmental leaps, separation anxiety, and teething, and the same boring-routine rule holds (HealthyChildren.org). If a log helps you spot which changes stuck, keep it to one tap so it does not cost sleep of its own.

How do 4-month regressions differ from 15-, 18-, and 20-month sleep regressions?

The 4-month event is not really a regression at all. Your baby's sleep cycles are maturing into adult-like architecture, and that change is permanent. Toddler regressions are different: they come from something you can usually name.

Toddler regressions: milestones, separation, molars

At 15 months, walking and language leaps disrupt sleep. At 18 months, separation anxiety peaks and naps start fighting back. At 20 months, molars and potty-training readiness add physical discomfort. Each has a trigger; the 4-month change does not, and it never "ends" the way a toddler phase does.

Comparison table: age, cause, typical duration

Age	Main cause	Typical duration
4 months	Sleep cycles mature permanently	2–6 weeks of rough nights, then a new baseline
15 months	Walking, language milestones	Days to ~2 weeks
18 months	Separation anxiety, nap resistance	~2–6 weeks
20 months	Molars, new independence	~2–3 weeks

The practical difference: after 4 months, the goal is adjusting to a permanent sleep structure. After 15, 18, or 20 months, you are waiting out a temporary storm that will pass on its own.

When should you talk to a pediatrician about regression sleep loss?

Most 4-month sleep disruption is normal, but a few symptoms fall outside a regression and deserve a call before the next well-child visit. The American Academy of Pediatrics advises contacting your pediatrician for feeding problems, fewer wet diapers, fever, unusual lethargy, or breathing pauses during sleep (healthychildren.org).

Symptoms that are not regression

Watch for: fewer than six wet diapers a day, refusing feeds, crying that cannot be soothed, snoring or pauses in breathing, and a baby who is hard to wake for feeds. Also flag sleep loss paired with poor weight gain. Regressions affect sleep, not growth or feeding.

What to bring to the appointment

Bring a doctor-ready PDF of your logs: feed times, sleep stretches, and wet diapers over the last three to five days. Concrete numbers turn a vague "sleeping badly" into a diagnosis your pediatrician can act on.

How can you track the regression without making 3 a.m. harder?

During a rough week, logging has one job: capture the data your pediatrician will ask for at the next well-child visit, without waking you up more than the baby already has. That means one-handed, one-tap entries for night feeds, sleep stretches, and wet diapers, and nothing else competing for your attention.

What to log and what to skip

Log feeds (bottle or nursing side), sleep stretches, and wet diapers. Skip anything that takes more than a tap or two: detailed nap-quality notes, mood tags, or session timers you have to babysit. If an entry needs both hands, you will abandon it by night three.

Honest limits of tracking during a rough week

No tracker shortens the regression or improves sleep. A calm tool like Calm Baby Tracker covers feeds, sleep, and diapers with no ads, and exports a doctor-ready PDF, but it will not fix a maturing sleep cycle. If logging feels like a second job, log less and sleep more.

FAQ

Can the 4-month sleep regression last longer than 6 weeks?

Yes. Some babies take closer to 8 weeks to settle, and the timeline stretches when a baby relied on heavy rocking or nursing to fall asleep before the change hit. If your baby is still waking every 60-90 minutes six weeks in, treat that as a signal to examine sleep habits rather than keep waiting.

Do all babies go through the 4-month sleep regression?

Every baby's sleep cycles mature around 3 to 5 months, but the visible disruption varies widely. Some babies show a sudden collapse in sleep; others show a few rough nights or barely a blip. The underlying change is universal, the symptom is not, which is why some parents never notice anything happened.

Is the 4-month regression the same as the 4-month sleep 'progression' some people talk about?

They describe the same event with different framing. 'Regression' is the traditional term because sleep looks worse. 'Progression' reflects what is actually happening: your baby's brain is building adult-like sleep cycles, a genuine developmental step forward. The word choice does not change the timeline, the wakings, or what you do about them.

What if my baby was sleeping through the night before the regression hit?

That is common and makes the change feel harsher, but it does not mean anything went wrong. A baby sleeping long stretches was riding shallow newborn sleep. Once mature cycles arrive, those stretches break up, and the skill of linking cycles has to be learned. Expect the old pattern to return in a new, matured form.

Does the regression come back after it ends?

The 4-month change itself does not repeat; once sleep cycles mature, they stay matured. But rough sleep returns at other ages: regressions around 15, 18, and 20 months, driven by milestones, separation anxiety, and teething. Those are temporary storms with identifiable causes, unlike the 4-month event, which is a permanent structural shift.