- 01A permanent change in sleep architecture, not a temporary phase — can start 3–5 months
- 02Typically lasts 2–6 weeks; can persist longer if sleep habits don't change
- 03Naps are hit hardest — the "45-minute nap trap" is the hallmark sign
- 048 tips: swaddle transition, drowsy-but-awake practice, protected wake windows, earlier bedtime, white noise, consistent routine, no new sleep crutches, tracking
- 051–3 night feeds are still normal at 4 months — the regression isn't purely hunger-driven
- 06Better sleep is built on this new sleep architecture, once babies learn to self-settle
What Is the 4-Month Sleep Regression?
This is not like the 8-month, 12-month, or 18-month regressions — those are temporary disruptions. The 4-month sleep regression is a permanent architectural change in how your baby sleeps. Sleep won't revert to the newborn pattern — but it can improve significantly once babies learn to self-settle between cycles.
The regression itself isn't the problem. The problem is that your baby may not yet know how to fall back to sleep independently at the end of a cycle. Some babies sail through with minimal disruption — "unicorn babies," as some sleep consultants call them — but the underlying sleep change is universal. Symptom severity simply varies from baby to baby.
Why the 4-Month Sleep Regression Happens: The Science
Newborn sleep has only two states: Active Sleep (similar to REM) and Quiet Sleep. Babies cycle between these roughly every 50 minutes, and when they wake at the end of a cycle, they're usually still drowsy enough to drift back without fully rousing [1].
At 3.5–5 months, the brain reorganises into adult-like four-stage sleep architecture: NREM Stage 1 (light), NREM Stage 2 (light), NREM Stage 3 (deep, slow-wave), and REM. Cycle length stays roughly 45 minutes [8]. The key change is that babies now wake fully at the transition between cycles — just as adults do, except adults already know how to fall back asleep. Babies who were nursed, rocked, or held to sleep haven't learned that skill yet.
At each wake point (●), your baby briefly surfaces. If they fell asleep in your arms, they'll look for that again.
Signs of the 4-Month Sleep Regression
Suddenly Waking More at Night
A baby doing 3–4 hour stretches may now wake every 45–90 minutes. This is the most reported sign.
Short Naps — the "45-Minute Trap"
Naps that cut off exactly at 45 minutes are a hallmark — that's the length of one sleep cycle.
Difficulty Settling to Sleep
A baby who was easy to put down now fights it — more aware of their environment at each cycle's end.
Increased Fussiness & Overtiredness
Broken sleep accumulates. By the afternoon, a baby who hasn't napped well may be extremely difficult to settle.
Wanting to Feed More at Night
Not usually hunger-driven — the breast or bottle is the sleep association they're reaching for.
Fighting the Swaddle
At 3.5–5 months, many babies begin rolling. The swaddle that helped them sleep may now be a hindrance — and a safety risk.
Shorter Total Sleep Despite More Effort
You're putting more in; they're sleeping less overall.
Developmental Signs Alongside
Rolling practice, increased hand interest, more visual tracking — milestones and this regression arrive together.
When Does the 4-Month Sleep Regression Start and End?
When Does the 4-Month Sleep Regression Start?
Not always at 4 months. The "4-month regression" is named for when it most commonly peaks — but it can begin as early as 3 months (12–13 weeks) and as late as 5 months (20 weeks). The 16-week sleep regression is the same event.
| When it hits | Age in weeks | What's typically happening |
|---|---|---|
| Early onset | 12–14 weeks (3 months) | Light sleep disruption beginning; not yet the full regression |
| Peak onset | 15–17 weeks (3.5–4 months) | Most families hit the full regression here |
| Late onset | 18–21 weeks (4.5–5 months) | Still completely normal; some babies restructure later |
| Note: "4 months" | 16–18 weeks | The regression is named for age, not an exact week |
How Long Does the 4-Month Sleep Regression Last?
This is the most-searched question about the regression — so here's the complete answer. Typical duration is 2 to 6 weeks. With no sleep habit changes, it can persist 3 to 6 months — not because the regression itself continues, but because the underlying issue (needing help to fall back asleep) remains. With consistent independent sleep practice, most families see significant improvement within 3–4 weeks. Importantly: the regression ends when the brain finishes reorganising, but the waking continues until the baby learns to self-settle.
Signs the 4-Month Sleep Regression Is Over
- Naps lengthen back beyond 45 minutes, or baby begins linking cycles independently
- Night wake frequency reduces — moving back toward 1–2 night feeds rather than 5–6 wakings
- Easier to settle at bedtime — baby accepts being put down drowsy
- Less fussiness by the afternoon — less accumulated sleep deficit
- Baby begins to stir at cycle end and resettles without a full wake-up
8 Tips to Get Through the 4-Month Sleep Regression
🌙 Getting Through It — 8 Tips
See the trend before the next sleepless night.
Log every nap and night wake, and Lunara's AI shows you whether waking frequency is actually reducing — not just whether tonight felt hard.
Track Your Baby's Sleep Free →The 4-Month Sleep Regression and Naps
Naps are often hit hardest. They involve short sleep periods — exactly one or two sleep cycles. When a 4-month-old reaches the end of a cycle (around 45 minutes), they wake fully. Before the regression, newborns often drifted back into the next cycle. Now they don't.
The 45-minute nap trap: if your baby consistently wakes at the 35–45 minute mark, this is almost certainly a cycle-end wake, not a hunger or discomfort wake. The test: go in before the 45-minute mark (around 35 minutes) and try to resettle before they fully wake.
| Time | Nap | Wake window before | Duration goal |
|---|---|---|---|
| ~7:00 AM | Wake for the day | — | — |
| ~8:30–9:00 AM | Nap 1 | 1.5–1.75 hours | 45–90 minutes |
| ~10:30–11:00 AM | Nap 2 | 1.5–1.75 hours | 45–90 minutes |
| ~1:00–1:30 PM | Nap 3 | 1.5–2 hours | 45–90 minutes |
| ~3:30–4:00 PM | Nap 4 (catnap) | 1.5–2 hours | 30–45 minutes |
| ~6:30–7:00 PM | Bedtime | 1.75–2 hours after catnap | — |
4-month-olds typically need 4 naps per day. The total daytime sleep target is 4–5 hours. If naps are running short (45 min each), you may need to attempt a resettle at the 35-minute mark, or accept more frequent naps and compensate with a slightly earlier bedtime.
The 4-Month Sleep Regression and Night Feeding
At 4 months, most babies still need 1–3 night feeds. Night waking during the regression is not purely hunger-driven — but that doesn't mean you withhold feeds. The distinction: feed at night because your baby is hungry; don't feed to settle back to sleep as the only tool you have.
Signs it's hunger: baby feeds actively for 5+ minutes, isn't easily resettled with other means, and it's been 3+ hours since the last feed. Signs it's comfort-seeking: baby settles quickly after only 2–3 minutes, falls asleep immediately when offered, and wakes again 45 minutes later. Feeding frequency at 4 months averages 1–2 per night in breastfed infants, less in formula-fed infants [3].
"Crying at night" — both hunger AND sleep associations can cause this, address both.
"Should I feed?" — yes to hunger; practice drowsy-but-awake for comfort wakes where you can.
"Night feeding" — 1–2 night feeds are still normal at 4 months, not a problem to solve.
4-Month Sleep Schedule: How Much Sleep Does a 4-Month-Old Need?
| Age | Total sleep | Night sleep | Naps | Wake windows |
|---|---|---|---|---|
| 3.5–4 months | 14–16 hours | 9–11 hours | 4–5 naps, 4–5 hours total | 1.25–1.75 hours |
| 4–5 months | 13–15 hours | 9–11 hours | 3–4 naps, 4–5 hours total | 1.5–2 hours |
These are averages. Some 4-month-olds sleep 13 hours total; some sleep 16. Track your baby's actual numbers rather than comparing to averages — Lunara shows you how your baby's sleep compares to the age-appropriate range. The AAP recommends 12–16 hours of total sleep, including naps, for infants 4–12 months [6].
4-Month Sleep Regression vs. Teething vs. Growth Spurt
| 4-Month Regression | Teething | Growth Spurt | |
|---|---|---|---|
| Age | 3.5–5 months | Usually 4–7 months+ (first tooth) | Any age — common at 3, 6, 9 months |
| Night waking pattern | Every 45–90 min (cycle length) | Irregular, often linked to pain timing | Every 2–3 hours (hunger-driven) |
| Naps affected? | Yes — short naps (45 min) | Sometimes — pain may interrupt | Often — hunger wakes cut naps |
| Key indicator | Waking at precise intervals | Drooling, gum inflammation, temp ≤ 38°C | Feeding more aggressively, weight gain |
| Lasts | 2–6 weeks | Until tooth erupts (days to weeks per tooth) | 2–3 days typically |
| Feeds at night | Comfort-seeking | Can increase (gum relief) | Hunger-driven — active feeds |
| Daytime behaviour | May be fine during the day | Often fussy/biting during day | Hungrier during day too |
These three conditions frequently overlap at 4 months. A baby can be teething, hitting a growth spurt, AND going through the sleep regression simultaneously — you don't have to pick one.
Is the 4-Month Sleep Regression Getting Worse?
If the regression is in week 6+ and shows no sign of improving, it's likely no longer the regression itself — it's a sleep habit issue the regression revealed.
- Waking 8–10+ times per night past week 4 — the regression has exposed a strong sleep association that needs addressing
- Naps haven't lengthened at all after 6 weeks — wake windows or the nap environment need adjustment
- Baby only sleeps when held and this is new — may be a separation anxiety sign coming in early
- You haven't changed anything in 4 weeks — the regression alone doesn't last this long; the sleep habits do
What to do:
- Choose and consistently apply one independent sleep practice — drowsy-but-awake at one nap daily
- Track for 7+ days in Lunara — if the AI shows no improvement trend over 2 weeks of consistent practice, it may be time to consult a paediatric sleep specialist
- Re-evaluate wake windows — overtiredness commonly extends regression symptoms
Common Mistakes During the 4-Month Sleep Regression
🚫 Keeping the Swaddle Past Rolling Readiness
Once baby shows signs of rolling, continuing the swaddle is a SIDS risk. The regression is hard enough — don't delay the transition until you're also managing a safety risk.
🚫 Moving to a Later Bedtime to "Tire Them Out"
This is the single most common mistake. Overtired babies sleep worse. Move bedtime earlier, not later.
🚫 Dropping the Bedtime Routine Because It "Isn't Working"
The routine works. The regression makes it look like it isn't. Keep it unchanged for consistency.
🚫 Assuming It's Hunger and Adding Extra Feeds All Night
Some night feeds are appropriate; responding to every waking with a feed reinforces the bottle or breast as the only sleep association.
🚫 Waiting for the Regression to Pass Before Working on Independent Sleep
The regression doesn't end because the baby naturally learns to self-settle. It ends when the brain finishes reorganising — but waking continues until the self-settling skill develops. These are separate. Start one daily practice now.
🚫 Comparing Your Baby to Other People's Babies
The range of "normal" at 4 months is enormous. A baby sleeping through the night is not ahead. A baby waking 5 times is not behind. The skill develops on different timelines.
Can You Sleep Train During the 4-Month Sleep Regression?
Short answer: yes, gentle sleep habit work can begin at 4 months. Full extinction methods ("cry it out") are a family decision but are not recommended before 4–6 months by most paediatric sleep associations.
A more useful framing: "sleep training" and "responding to the regression" are not opposites. You can practice drowsy-but-awake at one nap per day (not sleep training — just building a skill), lengthen your response time slightly at night before going in (graduated response), and begin to differentiate hunger cries from comfort or habit cries.
Some families choose cry-it-out from 4 months. It is not harmful according to available evidence [9][10]. It is also not required. Most important: consistency matters more than method. Pick one approach — switching approaches nightly teaches nothing.
How Lunara Helps You Track and Survive the 4-Month Regression
The regression feels endless when you can't see the trend. Lunara's sleep log gives you the data to know whether things are improving — or whether the regression has passed and something else is causing the wakings.
"Based on Arlo's sleep logs, night waking frequency has dropped from 5 per night to 2 per night over the past 7 days. His bedtime extension from 7:00 PM to 6:40 PM appears to be helping."
— Lunara AI Insight (illustrative example)Log every nap and night wake. After 7 days, Lunara's AI compares your baby's sleep to age-appropriate targets and surfaces patterns you can't see at 3am. Both parents log in real time — your partner's 2am log and your 5am log become one continuous record, so neither of you is guessing what happened overnight.
You can track your baby's sleep with Lunara, log night wakings with your partner in real time, and revisit building a sleep routine for your infant once the regression eases. Parents of twins can see navigating the 4-month regression with twins for a dedicated breakdown.
Track Your Baby's Sleep Free →
7-day free trial · No credit card required
Frequently Asked Questions
The 4-month sleep regression typically lasts 2 to 6 weeks. However, the duration depends significantly on whether babies develop the ability to fall back to sleep independently between sleep cycles. Without any change in sleep habits, the waking pattern introduced by the regression can persist for 3 to 6 months — not because the regression continues, but because the underlying sleep association remains. With consistent practice of independent sleep skills (such as drowsy-but-awake), most families see meaningful improvement within 3 to 4 weeks.
Yes. The "4-month" regression is named for when it most commonly peaks, but it can begin as early as 12 to 14 weeks (3 months) and as late as 20 to 21 weeks (around 5 months). The 16-week sleep regression is the same event. If your 3-month-old is suddenly waking more frequently at night after previously sleeping better, the regression may be beginning early.
The most common signs are: frequent night waking (every 45 to 90 minutes), naps cutting off at exactly 45 minutes, increased difficulty settling to sleep, greater fussiness due to accumulated tiredness, more frequent night feeding, and fighting the swaddle as rolling awareness develops. Some babies also show increased daytime feeding and seem more alert or stimulation-seeking than before.
Eight evidence-based tips: (1) Begin transitioning away from the swaddle if rolling is emerging. (2) Practice putting baby down drowsy-but-awake at one nap per day. (3) Protect age-appropriate wake windows (1.5–2 hours at 4 months). (4) Move bedtime earlier rather than later. (5) Use consistent broadband white noise. (6) Maintain the bedtime routine unchanged. (7) Avoid creating new sleep associations you will need to undo. (8) Track every sleep in an app to see improvement trends that are invisible night-to-night.
Yes — often more dramatically than nights. At 4 months, babies sleep in single sleep cycles of roughly 45 minutes during naps. When they wake at the end of a cycle, they may fully rouse rather than drift back into the next cycle. This results in naps that cut off precisely at the 35 to 45 minute mark. Attempting a resettle before the 45-minute mark (at around 30 to 35 minutes) can help bridge cycles. A 4-month nap schedule typically includes 4 naps per day totalling 4 to 5 hours of daytime sleep.
They are related but different. The regression refers to the disruption caused by the sleep architecture change. The schedule change refers to the shift in wake windows, nap counts, and bedtime that happens around the same age. Most 4-month-olds need to move from 4 to 5 naps toward 4 naps with slightly longer wake windows (1.5 to 2 hours). Both often happen simultaneously and are caused by the same underlying developmental shift.
No. The 4-month sleep regression introduces a permanent change in sleep architecture — but this does not mean permanent sleep disruption. Once babies learn to self-settle between sleep cycles (either naturally or with consistent practice), most sleep significantly better than they did as newborns. Many families find that their babies actually sleep better at 6 to 8 months than they ever did as newborns, once the regression has passed and independent sleep skills have developed.
This is very common during the regression and is typically driven by the comfort of the contact sleep association. Start with one practice session per day (not all day) of putting baby down drowsy-but-awake on a firm, flat sleep surface with white noise. Expect protest the first several attempts — this is normal. Consistency over 7 to 10 days is more effective than doing it every nap for 2 days and giving up. Track the sessions in Lunara to see progress you can't perceive night-to-night.
Related Articles
References
- Blumberg MS, et al. (2014). Spatiotemporal structure of REM sleep twitching. Curr Biol.
- Mindell JA, et al. (2009). A nightly bedtime routine: impact on sleep in young children. Sleep.
- Henderson JMT, et al. (2011). Sleeping through the night: the consolidation of self-regulated sleep. Pediatrics.
- Galland BC, et al. (2012). Normal sleep patterns in infants and children. Sleep Med Rev.
- Tham EKH, et al. (2017). Infant sleep and its relation with cognition and growth. Sleep Med Rev.
- American Academy of Pediatrics. (2022). Safe Sleep Recommendations.
- American Academy of Sleep Medicine. (2016). Recommended amount of sleep for paediatric populations.
- Bathory E, Tomopoulos S. (2017). Sleep regulation, physiology, and development. Curr Probl Pediatr Adolesc Health Care.
- Sadeh A, et al. (2010). The role of actigraphy in the evaluation of sleep disorders. J Clin Sleep Med.
- Owens JA. (2004). Sleep in children: cross-cultural perspectives. Sleep Biol Rhythms.
See the trend before the next sleepless night.
Lunara tracks every nap and night wake — and shows you when the regression is actually ending.
Track Your Baby's Sleep Free →