Every list of baby sleep regression ages and stages you will find online is built on a word that points the wrong way. "Regression" means a skill was present and then lost — that sleep worked, and then broke. That is almost never what happened. At each of these ages the baby moved forward: the electrical structure of sleep matured, a new motor skill arrived, the understanding that you still exist after you leave the room switched on, or the amount of daytime sleep the routine was built around quietly stopped fitting. Nothing went backward. Something advanced, and last month's schedule no longer covers this month's baby.
That sounds like a semantic quibble. It is not, because the two framings send you to opposite responses. If you believe sleep broke, you hunt for a fix — a stricter method, a new product, a harder line at 2am. If you understand that your baby's needs shifted underneath a schedule that stayed put, you go looking for what moved and adjust it. The second approach resolves far more of these nights, and it asks almost nothing of you in the dark.
There is a second problem with the standard list. The ages people recite — four months, eight to ten months, twelve months, eighteen months — are not variations on one phenomenon. They are genuinely different events flattened under one label. Four months is a permanent, one-way change in how sleep is built. Eight to ten months is motor development plus separation awareness. Twelve and eighteen months are usually nap-transition problems wearing a costume. Treating them as one thing implies they all pass the same way and answer to the same response. They do not, and expecting them to is what keeps families stuck.
What follows maps each age to what is actually maturing, how long the disruption tends to run, and — the part that matters most — whether you should wait it out or change the schedule. Where the evidence is thin, and for several of these ages it is genuinely thin, this page says so rather than inventing confidence.
Need the clock times for your baby?
The calculator builds the whole day from one wake-up.
Start by throwing out the word "regression"
There is no clinical entity called a sleep regression. It does not appear in American Academy of Pediatrics sleep guidance, in the American Academy of Sleep Medicine literature, or in any pediatric sleep textbook as a diagnosis. It is a parenting-forum term, and like most parenting-forum terms it started as useful shorthand and quietly picked up baggage.
The baggage is the implied direction of travel. "Regression" tells a depleted parent that something was lost, that they are back at square one, and — usually unspoken, always felt — that they may have caused it. Parents ask whether they built a bad habit or undid their progress. That framing is not merely demoralising; it aims you at the wrong lever. It pushes you toward tightening your night response when the actual answer is usually loosening or rebuilding the daytime structure.
Reorganization is the more honest word. At each of these ages something in the system genuinely changes: the electrophysiology of sleep, the way sleep distributes across twenty-four hours, the baby's motor repertoire, or their mental model of where you go. Sleep does not get worse in any absolute sense. It gets different, and the settling routine that fit last month's baby stops fitting.
The distinction that actually earns its keep is developmental versus logistical. A developmental change — matured sleep architecture, the arrival of object permanence — is not something you fix. You accommodate it, and it eases as the new capacity consolidates. A logistical mismatch — too much day sleep, a wake window that outgrew the schedule, a nap the child no longer needs — will never resolve on its own, because nothing matures its way out of an arithmetic error. Most of the misery in these months comes from applying the wait-it-out response to a logistical problem, or the change-everything response to a developmental one.
Read the four-month change as permanent, not a phase
Of every age on the list, this one is categorically different, and misreading it costs families months.
Newborn sleep is not built like adult sleep. In the early weeks infants alternate between active sleep and quiet sleep — precursors to REM and non-REM that do not yet show staged architecture. Newborns also enter sleep through active sleep, which is a large part of why a three-week-old can be fed to sleep, carried across a room, set down, and stay under, while a five-month-old absolutely cannot.
Somewhere around two to four months that reorganizes. In the review of early childhood sleep architecture by Lenehan and colleagues, active sleep becomes REM and quiet sleep becomes NREM after about two months, and the distinct NREM stages begin to appear — sleep spindles are expected on EEG by around three months. Sleep onset migrates too: instead of dropping straight into an active state, the baby now begins sleep in light NREM, and by six months essentially all sleep should begin that way. Cycles stay short by adult standards, commonly described as roughly fifty minutes against an adult's ninety, so a baby surfaces toward waking many times a night. Underneath it, the circadian system consolidates and a real day–night rhythm establishes.
The practical consequence is blunt. Your baby now notices. They surface at a cycle boundary, find the room different from where they fell asleep, and call for the original conditions to be restored. Nothing broke — perception improved.
This is why four months does not "pass" the way the later ages do. The acute stretch of frequent waking and thirty-minute naps usually settles within two to four weeks. The architecture never reverts. Families waiting for sleep to return to what it was in month two are waiting for something that is not coming, and that wait is where the despair lives. What helps is not a trick: it is moving the falling-asleep conditions toward something still true at 2am. Our wake windows for a 4-month-old page has the daytime version, and the homepage has the whole-picture map.
- Sleep onset shifts from active sleep to light NREM — the transfer from arms to crib stops working
- Distinct NREM stages emerge, so surfacing between cycles becomes a real waking event
- Cycles run roughly 50 minutes, meaning many nightly opportunities to notice a changed room
- Circadian rhythm consolidates, which is why night sleep finally starts clumping together
Expect eight to ten months to be motor, not behavioral
This window is where physical development and cognitive development collide, and it produces the strangest night behavior of the first year.
Babies rehearse new motor skills in their sleep. A baby who learned to pull to stand on Tuesday will pull to stand in the crib at 1am on Wednesday, asleep in every sense that matters, then wake genuinely distressed because they are upright with no idea how to get down. This is not defiance and it is not a habit forming. It is consolidation — a nervous system running drills. It burns itself out in days to a couple of weeks per skill, and the highest-yield intervention is entirely a daytime one: drill the reverse of the skill on the floor. Practice sitting down from standing until it is automatic, and the midnight version usually stops shortly after.
Running beneath the motor work is a cognitive shift. Object permanence matures across the second half of the first year. Before it, you leaving the room is neutral information. After it, you leaving means you exist somewhere else and the baby would very much like to be there. Separation protest at bedtime and on night wakings commonly sharpens here for exactly that reason. It is a milestone, not a setback, and it is not evidence that anything you did stopped working.
The third element is logistical and it is the one most often missed: many babies are ready to move from three naps to two somewhere in this window. If the third nap is being fought, is shoving bedtime past 8pm, or only happens in a moving car, that nap is finished. Holding it produces a late bedtime and, counterintuitively, a worse night.
- Drill the reverse of each new skill on the floor — sitting down from standing, lying down from sitting
- Give generous unrestricted floor time; a baby who moves all day rehearses less at midnight
- Keep the night response short, dull and predictable — this window rewards consistency over cleverness
- Audit the third nap: a fight, or a bedtime past 8pm, usually means it is done
- Practice cheerful five-second separations by day so the pattern that you come back gets built in daylight
So why do twelve and eighteen months feel identical?
Because they usually share one cause, and it is not a mysterious leap. It is the road from two naps to one.
Here is the trap. Around twelve months a great many babies start refusing the morning nap, or taking it and then refusing the afternoon one. This reads unmistakably as ready for one nap. For most twelve-month-olds it is not. It is far more often a wake-window problem: the morning window has grown, the first nap is now offered too early, and the baby simply is not tired enough to take it. Pushing that nap later — often to somewhere between 9:30 and 10am — frequently rescues the two-nap day and buys several more months of it.
Dropping to one nap prematurely backfires in a recognisable pattern: overtiredness by mid-afternoon, bedtime becomes a fight, night waking climbs, and early-morning waking appears. If those four arrived together shortly after you dropped a nap, the nap drop is your explanation, and it is reversible.
The genuine transition lands for most children between roughly fourteen and eighteen months, and it is legitimately brutal, because there is a stretch of weeks where one nap is not quite enough and two no longer fit. Our 2 to 1 nap transition page covers running it, and the 3 to 2 nap transition page handles the earlier one.
Eighteen months layers autonomy on top. This is the age a toddler discovers bedtime is a place where they hold leverage. Stalling, requests, climbing out, and one more book are developmentally on time. They are a boundary question rather than a sleep question, and they answer to an unhurried, non-negotiable routine far better than to any tinkering with sleep timing.
- Before dropping a nap, try pushing the morning nap 30 minutes later for a week
- Four symptoms arriving together after a nap drop — overtiredness, bedtime fights, night waking, early rising — means go back
- Expect the real 2-to-1 transition between 14 and 18 months, not at 12
- Treat 18-month bedtime stalling as a limits problem, not a sleep-need problem
Check three things before you call it developmental
The single most useful skill here is telling a genuine reorganization apart from the three impostors: illness, teething, and plain schedule drift. They present similarly at 3am and respond to completely different things.
The first check is onset shape. Developmental disruption tends to arrive alongside a visible new skill and builds over several days. Illness arrives abruptly, often overnight, and usually drags something else with it — a temperature, congestion, a changed cry, refused feeds, fewer wet nappies. Schedule drift arrives gradually and from the wrong end of the night: early-morning waking and shrinking naps first, night waking later.
The second is the daytime read. A baby in a developmental window is typically happy and busy by day, rehearsing constantly, feeding normally. A sick baby is off in daylight too. This one check resolves a surprising share of cases.
The third is paper. Write down actual clock times for four consecutive days — wake, every nap start and end, bedtime, every night waking. Schedule causes become obvious on paper in a way they never do in your head at 5am, and the exercise takes four minutes a day. If total daily sleep has crept outside the age range, or the last nap is ending too close to bedtime, you have found it.
Teething deserves particular scepticism. Teeth arrive across roughly two years, so something is always erupting and teething is always available as an explanation — which makes it a weak one. The measured effects are real but mild and clustered tightly around the day or two of actual eruption, not spread across three weeks.
- Log four days of real clock times before you conclude anything
- Ask whether the daytime baby is also off — if yes, think illness first
- Sustained disruption with no new skill and no symptoms points at the schedule
- Congestion, noisy breathing, or pauses in breathing during sleep go to your pediatrician, not to a sleep method
| Signal | Developmental shift | Illness | Schedule drift |
|---|---|---|---|
| How it starts | Builds over days, alongside a new skill | Abrupt, often overnight | Gradual, over two or three weeks |
| Daytime mood | Normal — busy, rehearsing, feeding well | Clearly off; feeding may drop | Cranky late in the day, fine early |
| First symptom | Night waking or crib practice | Fever, congestion, changed cry | Early rising and shortening naps |
| What resolves it | Time plus a consistent response | Treating the illness | Changing wake windows or dropping a nap |
Hold one response steady and protect the daily total
Very little meaningfully shortens a genuine reorganization. Quite a lot stops you accidentally extending it or building a second problem that outlives the first.
The most common way families stretch a rough patch is changing the response nightly. Night one you feed, night two you rock, night three the baby comes into your bed, night four you are back to feeding. Every one of those is a defensible choice. Rotating between them at random is reliably harder on a baby than any single one of them, because there is no pattern available to learn. Pick the response you are genuinely willing to sustain for a week and sustain it — not because consistency is virtuous, but because it is the only thing that produces information.
The second is compensating with day sleep. A savage night makes a long morning nap feel like mercy, and it is, right up until the extra daytime sleep comes straight out of the night and you are two weeks into a self-feeding cycle. Holding the daily total roughly where it was, and defending the timing of the last nap and bedtime, does more work than almost anything else available to you.
On totals: the American Academy of Sleep Medicine's pediatric consensus recommends 12 to 16 hours per 24 hours for infants 4 to 12 months including naps, and 11 to 14 hours for children 1 to 2 years. Those ranges are wide deliberately. A baby at the bottom is not deprived and one at the top is not lazy. Use the range to sanity-check a total, never to chase a number. And follow safe-sleep guidance throughout — flat, firm, empty surface, baby on the back, room-sharing without bed-sharing — which does not relax because everyone is tired.
- Choose one night response and keep it a full week before judging it
- Do not let day sleep expand to cover a bad night — defend the total and the bedtime
- Keep the room genuinely dark and the routine identical, especially when everything else is chaos
- Get outdoor light in the morning; daylight is the strongest input to a developing circadian clock
- Keep the sleep surface flat, firm and empty every night, tired or not
Realistic duration, and the six-week outer limit
Most published durations for the baby sleep regression ages and stages below come from parent report rather than measurement, so treat every number as a typical range and not a constant. The exception is four months, where the underlying architectural change is documented — the acute disruption passes, but the change itself is permanent.
Two caveats before the map. Individual variation is enormous: the systematic review of normal infant and child sleep by Galland and colleagues found wide spreads around every average at every age, which is the polite way of saying your baby may do none of this on schedule. And more than one thing usually runs at once. A ten-month-old may be pulling to stand, working through separation anxiety, and ready to drop a nap simultaneously. Untangling those is most of the work.
The last column is the one to read. "Wait" means the disruption is a byproduct of maturation and will ease as the capacity consolidates. "Adjust" means something structural is now wrong and will stay wrong until you change it.
The outer limit matters more than any single row. A genuine reorganization ends. If a disruption has run past six weeks with no sign of movement, drop the regression framing entirely and consider something else — reflux, an ear infection, a milk protein issue, disordered breathing in sleep, or a schedule that drifted and never came back. Persistent unexplained night waking, or noisy or laboured breathing during sleep, is a conversation with your pediatrician rather than something to outlast. That is not a hedge; it is the single most useful sentence on this page.
| Age | What is actually maturing | Typical duration | Wait or adjust? |
|---|---|---|---|
| 3–4 months | Sleep architecture and circadian rhythm; NREM stages emerge | 2–4 weeks acute — but the change itself is permanent | Both |
| 6 months | Often nothing. Usually teething, first illness, or solids | Days to 2 weeks, resolves with the trigger | Wait, and look for a cause |
| 8–10 months | Crawling, pulling to stand, object permanence; third nap ends | 2–6 weeks | Both — support the skill, drop to two naps |
| 12 months | Walking, language comprehension. Two naps usually still needed | 1–3 weeks | Adjust — but resist dropping a nap yet |
| 14–18 months | Genuine 2-to-1 nap transition, plus autonomy and molars | 3–6 weeks; often the hardest of the lot | Adjust — schedule is the main driver |
| 24 months | Imagination and first fears; total sleep need falls | 2–4 weeks, mostly bedtime resistance | Both — trim day sleep, hold the boundary |
Questions parents ask
Are sleep regressions real, or is the whole idea a myth?
The disruptions are entirely real. The label is imprecise. There is no clinical diagnosis called a sleep regression in AAP or AASM materials. What is real is that predictable things happen at predictable ages — architecture matures around three to four months, motor skills and separation awareness sharpen later in the first year, nap needs drop at intervals — and each disrupts a routine that was working. The experience is genuine; the word simply points backward when the baby went forward.
Which of these ages is hardest?
Parents most often name four months and the 2-to-1 nap transition around fourteen to eighteen months, for different reasons. Four months is hard because it is first, it lands when you are already depleted, and the change is permanent rather than temporary. The 2-to-1 transition is hard because there is a stretch of weeks where one nap genuinely is not enough and two genuinely no longer fit, so a small sleep debt runs while the day reorganizes. There is no way to skip that middle.
How long does one actually last?
Across the baby sleep regression ages and stages, two to six weeks covers most of them. The four-month change usually settles acutely in two to four weeks, while nap-transition disruptions run longer — six weeks or more is common, because the schedule change genuinely takes that long to bed in. These are parent-reported ranges, not measured constants, so hold them loosely. The more useful number is the outer limit: past six weeks with no improvement, stop treating it as developmental.
Can you prevent them?
Not the developmental parts. Sleep architecture will mature and your baby will learn to pull to stand regardless of anything you do. What you can prevent is the schedule drift that makes those changes far worse than they need to be. Keeping wake windows appropriate for the age, protecting the daily total, stopping bedtime creeping later, and adjusting naps promptly all reduce the damage. Roughly half of what gets called a regression is a schedule that stopped fitting — and that half is preventable.
Is there really a six-month sleep regression?
It is the weakest entry on the standard list. Six months is not associated with a distinct maturational shift the way three to four months is, and it does not usually coincide with a nap transition. When sleep genuinely falls apart at six months the cause is more often identifiable — teething, a first illness as maternal antibodies wane, starting solids, or the aftermath of travel. That is good news: a specific cause is something you can address, unlike a phase you can only outlast.
My baby slept through the night and stopped. Did I cause it?
Almost certainly not. Consolidated night sleep in the first year is not a one-way achievement. Weinraub and colleagues, tracking infants from six to thirty-six months, identified distinct trajectories rather than steady progress, with a substantial group continuing to wake regularly well past the point their peers had stopped. A run of good nights followed by a run of bad ones is a normal shape for the first year, not evidence that a habit formed or a window was missed.
Should I sleep train during one?
Most guidance suggests waiting for the acute phase to pass, for practical rather than moral reasons. During a genuine developmental disruption your baby's capacity has actually changed, so any approach you begin will look like it is failing when it is really just being applied at the worst possible moment — which usually means you abandon it and conclude it does not work. Waiting two to three weeks gives you a far cleaner read. Our page on how to sleep train a baby covers the approaches.
Does teething explain any of this?
Less than it gets credit for. Teeth arrive across roughly a two-year span, so something is always erupting and teething is always an available explanation — which makes it a poor one. Research on teething symptoms generally finds effects that are real but mild and short-lived, clustered around the day or two of actual eruption rather than spread over weeks. If sleep has been disrupted for three weeks, teething is not the whole story. Check the schedule before blaming the molars.
What if my baby never seems to have one?
Then nothing is wrong. These ages describe population tendencies, not requirements, and the normative literature finds very wide variation around every average. Some babies pass through the four-month architecture shift with a handful of unsettled nights. Some move between nap schedules without visible protest. A baby who does not have a dramatic disruption at four or eight or eighteen months is not behind, is not skipping a stage, and is not storing it up for later.
Do these show up in naps as well as at night?
Yes, and naps often go first. Daytime sleep has less homeostatic pressure behind it than night sleep, so it is the more fragile of the two and tends to register a change earlier. Short naps appearing before night waking is a common opening sequence, particularly around four months and during nap transitions. If naps shortened first and nights followed a week or two later, that ordering itself is a useful clue that the schedule, not the baby, is what moved.
Is a sudden 45-minute nap at four months part of this?
Usually, yes. A forty-five-minute nap is roughly one infant sleep cycle, and it appears at this age for the same reason the nights fall apart: the baby now surfaces at the cycle boundary and notices the room. It is not a sign that naps are broken. Most families see nap length rebuild over the following weeks as the transition between cycles becomes something the baby can ride through, though it rarely returns to newborn-style four-hour stretches.
Does a sleep sack help through these stages?
It will not shorten a developmental change, and anything promising that is overselling. What a sleep sack does is make one variable of the night identical every time, which is genuinely useful when a baby has just started noticing that things differ between falling asleep and waking. It also keeps the surface free of loose bedding, which matters more once a baby is mobile enough to move around the crib. It was the one item I actually relied on — as a constant, not a cure.
Sources
- Lenehan et al., The Architecture of Early Childhood Sleep Over the First Two Years (Maternal and Child Health Journal, 2023)
- AASM — Recommended Amount of Sleep for Pediatric Populations (consensus statement)
- Galland et al., Normal sleep patterns in infants and children: a systematic review (Sleep Medicine Reviews, 2012)
- Weinraub et al., Patterns of developmental change in infants' nighttime sleep awakenings from 6 through 36 months (Developmental Psychology, 2012)
- Sleep Foundation — Infant Sleep Cycles
- Cleveland Clinic — Sleep in Your Baby's First Year
- NICHD Safe to Sleep
Know a tired parent who needs this?
Send it their way — it is free and always will be.
Keep reading
Building a full day from scratch? Start with the baby sleep schedule guide and calculator on the homepage.
Written by
Sukie Gao
Mom of a daughter, and the person behind every tool on this site.
Sukie Gao has a daughter, and when her daughter was a newborn she was exhausted and completely out of her depth. She spent months piecing together sleep advice from scattered corners of the internet, and built My Baby Sleep Schedule so the next tired mother would not have to search as hard. She writes as a parent from lived experience, not as a clinician.
Last reviewed and updated . Sukie is a parent, not a clinician — check anything that matters with your pediatrician.