A sleep regression is a temporary period, usually lasting 2 to 6 weeks, when a previously well-sleeping baby or toddler suddenly starts fighting sleep, waking more often, taking shorter naps, or refusing bedtime. Regressions happen at predictable developmental stages: 4 months, 6 months, 8 to 10 months, 12 months, 18 months, and 2 years. Each is driven by different developmental leaps but responds to a similar universal playbook: hold the schedule, hold the routine, do not introduce new sleep crutches, and give your baby daytime practice for the new skill their brain is working on.
What sleep regressions actually are
A sleep regression is one of the most confusing experiences of early parenthood. Your baby was sleeping well. Nothing about your routine has changed. Your baby's environment is the same. Their diet is the same. And yet, from one night to the next, everything falls apart.
The technical definition of a sleep regression is a period of 2 to 6 weeks where a previously well-sleeping baby suddenly regresses in their sleep. In practice, this means one or more of the following happens:
- Nights that were solid start becoming broken again
- Wakings appear that had disappeared
- Naps that were long become 30 to 45 minutes
- Bedtime that was calm becomes a 45 minute battle
- Your baby suddenly needs your active presence to fall asleep or resettle after being able to self-settle previously
- Early morning wakings appear (4am to 5am starts becoming the new "morning")
The reason parents feel blindsided is that regressions almost never have an environmental cause. The trigger is inside your baby's brain, not in anything you did or changed. This is why the fixes are also mostly internal to your baby's development rather than in your routine.
Are sleep regressions even real?
This is worth addressing head-on because you will find both perspectives online.
The case that they are real: Predictable disruptions in infant sleep have been documented across large longitudinal studies. Multiple studies confirm that sleep tends to worsen around specific developmental stages, particularly 4 months and 8 to 10 months. The pattern is consistent enough that pediatric sleep researchers use the term "regression" or "sleep disruption at developmental transitions" in academic literature.
The case that they are overused: Some sleep specialists argue that the "regression" label gets applied to any period of bad sleep, including situations that are actually driven by schedule problems, environmental issues, or a sleep skill deficit that was masked when the baby was younger. In other words, some parents blame "the 4 month regression" for issues that are actually the baby not having strong independent sleep skills.
The honest middle ground: regressions at 4 months and 8 to 10 months are well-supported by both research and clinical observation. The 6 month, 12 month, 18 month, and 2 year regressions are more variable, more clearly tied to specific developmental milestones, and not every baby experiences them significantly.
The practical implication: if your baby's sleep suddenly gets worse and they are at or near one of the regression ages below, it is reasonable to work with the regression framework. If your baby's sleep has been persistently bad for months without a clear developmental link, the issue is probably not a regression and may be a schedule or sleep skill problem instead.
Why sleep regressions happen (the neuroscience)
The underlying driver of every regression is the same: your baby's brain is doing something significant, and sleep pays the price temporarily.
Three things happen in the brain during a developmental leap:
New neural connections form rapidly. The brain literally rewires itself during major milestones. This process consumes energy and disrupts existing patterns, including sleep patterns.
Skills consolidate during sleep. When your baby learns a new physical or cognitive skill during the day, the brain does the work of turning it into a lasting ability while your baby is asleep. Deep sleep is the consolidation phase. During intense skill acquisition, the brain interrupts sleep to practice, which is why babies literally practice crawling, standing, or new words in their cot at 2am.
Skill practice takes priority over rest. The developing brain is programmed to prioritise mastering new skills. If given a choice between deep sleep and practicing a new movement, the brain often chooses the practice. This is temporary and it is not something the baby has any control over.
Understanding this changes how you respond. The regression is not something to solve. It is something to support while it runs its course.
Signs your baby is in a sleep regression
Regressions have a distinctive pattern that separates them from other sleep issues.
- Sudden onset: sleep worsened dramatically from one week to the next
- Timing near a known regression age (4, 6, 8 to 10, 12, 18, or 24 months)
- Visible developmental changes happening at the same time (new physical skill, language leap, separation anxiety)
- Fussiness during the day that seems disproportionate
- Fighting sleep even when clearly tired
- Multiple new wake times overnight
- Naps shortening or being refused entirely
- The baby seems otherwise healthy and fed normally
- No environmental changes explain the shift
If you tick most of these boxes and your baby is at or near a regression age, you are almost certainly dealing with a regression.
The main sleep regressions by age
Here is each regression in detail. Every baby is different. Some sail through several, some are hit hard by one, some experience regressions slightly earlier or later than the standard ages.
4 month sleep regression
When it hits: 3 to 5 months
How long it lasts: Permanent (this one is technically not a regression at all)
What is happening: Around 16 weeks, your baby's sleep architecture undergoes a permanent neurological shift. Newborn sleep has just two stages: active and quiet. Adult sleep has four stages, including light and deep sleep with clear cycles. At 4 months, your baby's sleep becomes structurally adult-like, which means they start briefly surfacing between sleep cycles the way adults do.
Why parents feel it: Adults briefly wake between sleep cycles too but resettle without noticing. Babies at 4 months have not yet learned how to bridge these transitions. So the wakings that used to be invisible now become fully awake, often crying, wakings.
The catch: This is not a temporary regression. It is a permanent change in how your baby sleeps. What resolves is your baby's ability to bridge sleep cycles independently, not the underlying brain change.
We cover the 4 month sleep regression in detail in our dedicated guide on how to survive the 4 month sleep regression.
6 month sleep regression
When it hits: 5 to 7 months
How long it lasts: 1 to 3 weeks
What is happening: This one is contested. Some sleep experts consider it a real regression driven by the introduction of solid foods, teething, and early mobility skills. Others argue that most 6 month "regressions" are actually schedule issues (baby needs a new nap schedule as they drop the third nap) or the tail end of the 4 month adjustment.
Signs: Frequent night wakings, resistance to naps, teething discomfort, and the beginning of separation anxiety in some babies.
The most likely underlying cause: Nap transitions (3 to 2 naps) that have not been addressed, teething pain, or a wake window that has not been updated as your baby has grown.
8 to 10 month sleep regression
When it hits: 7 to 10 months
How long it lasts: 2 to 6 weeks
What is happening: Four things at once. Crawling and other mobility skills. Object permanence (the cognitive realisation that things and people exist even when out of sight). Separation anxiety. The 3 to 2 nap transition. All of these hit within the same fortnight for many babies, which is why this regression often feels harder than the 4 month one.
Signs: Multiple night wakings, standing up in the cot at night, fighting bedtime, refusing the third nap, crying when you leave the room, early morning wakes.
What actually helps: Extensive daytime practice for new physical skills, keeping bedtime and routine consistent, extending wake windows appropriately for the older age, and addressing separation anxiety with lots of daytime connection.
We cover this regression in detail in our dedicated guide on the 8 month sleep regression.
12 month sleep regression
When it hits: 11 to 13 months
How long it lasts: 2 to 4 weeks
What is happening: First steps often appear in this window, along with a burst of language development. Many babies also drop from 2 naps to 1 around this age, which creates a schedule disruption. Separation anxiety often intensifies with increased mobility (the baby can now leave you as well as be left by you).
Signs: Nap refusal, especially the morning nap. Night wakings. Bedtime resistance. Early morning wakes. Some babies also become clingy or fussy during the day.
What actually helps: Time. This one usually resolves faster than the 8 month one. Support the nap transition by dropping to 1 nap when your baby is ready (usually 13 to 16 months). Keep the schedule and routine steady during the transition weeks.
18 month sleep regression
When it hits: 17 to 19 months
How long it lasts: 2 to 6 weeks
What is happening: A vocabulary explosion. Toddlers at this age typically learn dozens of new words per week. This intense cognitive load disrupts sleep. On top of this, they are learning to assert independence, which shows up as bedtime resistance ("no!" becomes the answer to almost everything).
Signs: Sudden and strong bedtime resistance. Calling out from the cot repeatedly. Extended night wakings where the toddler wants to interact. Early morning wakes. Sometimes fully coming out of the cot for the first time.
What actually helps: Consistency and boundaries. This is the first regression where "the toddler is testing" becomes a real dynamic. Keep the routine short, calm, and firm. Do not introduce new negotiation patterns you do not want long term. Address any nap issues (some toddlers try to drop the last nap during this window, which almost always creates worse night sleep. Hold the nap for as long as possible).
2 year sleep regression
When it hits: 22 to 26 months
How long it lasts: 2 to 6 weeks
What is happening: Several possible drivers overlap in this window. A new sibling. Potty training starting. Transition from cot to toddler bed. Imagination developing (which can create fears of the dark or "monsters"). Language becoming complex enough for the toddler to argue back. Nightmares often appear for the first time.
Signs: Bedtime resistance, delaying tactics ("one more story", "I need water"), night wakings from bad dreams, fear of the dark, coming out of bed repeatedly, potentially early morning wakes.
What actually helps: Address the underlying trigger (new sibling adjustment, potty training pressure, etc). Delay the cot-to-bed transition if possible until the regression has passed. Introduce a nightlight if darkness is now scary. Keep boundaries firm but response warm. This regression is more about behaviour and emotion than pure development, so the response is more relational than schedule-based.
The universal playbook (what to do in ANY regression)
The specific triggers vary by age. The response is remarkably similar for every regression. If you internalise this playbook, you have the toolkit for every regression from 4 months to 2 years.
Hold the schedule
Whatever schedule was working before the regression, keep it. The brain craves consistency when everything else is changing. Do not shift bedtime, restructure naps, or try a new approach in the middle of the regression. Hold what you had. Even if it is not working perfectly right now, the underlying structure is more valuable than short-term relief.
The exception is age-appropriate adjustments (the 4 to 3 nap transition, the 3 to 2 nap transition, the 2 to 1 nap transition). If your baby's age is at one of these transitions AND the regression is hitting, the schedule may genuinely need to change. Our wake windows by age guide covers when each transition typically happens.
Hold the bedtime routine
Same order. Same timing. Same cues. Every night. If your routine was bath, feed, book, cuddle, cot, then that is the routine now too. Do not add new steps. Do not remove steps. Do not shorten it because you are tired and you just want them down faster.
The routine is a cue system. During a regression, the cues become more valuable, not less. They give your baby something predictable to anchor to while their brain is in flux.
Do not introduce new sleep crutches
This is the rule that decides whether the regression lasts 3 weeks or 3 months.
A baby waking at 2am screaming is the easiest baby in the world to feed back to sleep, rock back to sleep, or bring into your bed. Any of those work in the moment. The problem is that the baby will then expect the same response every night, long after the regression clears.
If your baby was settling independently before the regression, keep that approach now. If they were feeding once overnight, keep it at once, not three times. Do not increase the response just because the wakings have increased. The regression ends. The new habits you have created during the regression do not.
If you have already introduced sleep crutches during a regression and want to gently reverse them, our guides on how to break the feed-to-sleep association and how to gently break contact napping cover the reversals.
Support new skill practice during the day
Whatever skill your baby is working on during the regression (crawling, standing, walking, talking), give them extensive daytime practice. The more the skill gets consolidated during the day, the less the brain needs to schedule practice sessions at 2am.
For 8 month regression babies, this means floor time to practise crawling and pulling up. For 12 month regression toddlers, this means walking practice. For 18 month regression toddlers, this means lots of language input (books, conversation, songs). For 2 year regression toddlers, this often means addressing the underlying emotional or behavioural driver.
Adjust wake windows for the stage
Almost every regression is easier when wake windows are correct for the age. Babies who are on the wake windows that worked at their previous stage are often now undertired at bedtime, which makes settling harder.
At 4 months, wake windows are usually 90 to 120 minutes. At 8 months, 2.5 to 3.5 hours. At 12 months, 3 to 4 hours. At 18 months, 4 to 5 hours. Check your baby's wake windows against the current age, not the age they were 3 months ago.
Our wake windows by age guide covers the full chart.
Use tools that support without creating dependencies
The goal during a regression is to give your baby comfort and support without becoming the only source of that support.
Environmental supports (white noise, sleep sack, rhythmic patting from a tool) work better than parent-led supports (rocking, holding, feeding to sleep) during regressions because they do not require your active presence every time.
Our LullaBear™ provides gentle rhythmic patting that stays in the cot with your baby, offering the settling input without you having to be physically present at every waking. Our LullaHush™ portable white noise machine provides the continuous auditory environment that helps a baby bridge sleep cycle transitions. These are environmental sleep associations that persist through the regression and after.
For younger regressions (4 to 6 months) where the swaddle transition is often part of the difficulty, our LullaWrap™ provides the containment support many babies need during this stage.
When it is not a regression
Not every stretch of bad sleep is a regression. Before applying the regression framework, rule out these alternatives.
Teething. Teething discomfort disrupts sleep and can look like a regression. If your baby has visible red gums, is drooling more, chewing on everything, or has one or two nights of hard fussing rather than 2+ weeks of pattern change, teething is the more likely cause.
Illness. A cold, ear infection, urinary tract infection, or virus can disrupt sleep for days to weeks. If your baby has any signs of being unwell (temperature, congestion, changed appetite, unusual fussiness), rule out illness first.
Environmental changes. New house. New room. New caregiver. Daylight savings time. A change in routine due to a work schedule shift or a holiday. All of these can create regression-like patterns without being a regression.
Schedule problems. Wake windows that are out of date. Naps at the wrong time. Bedtime too late (or too early). Total daytime sleep off. These often look like regressions but are actually about the schedule needing an adjustment.
A sleep skill deficit. A baby who has always fed to sleep, been rocked to sleep, or contact napped has not built the skill of independent sleep. When they hit a stage where independent sleep becomes more important (like the 4 month cycle transitions), the lack of skill becomes visible. This can look like a regression but is actually an underlying issue that was masked previously.
Reflux or silent reflux. Persistent sleep issues in babies with feeding-related discomfort are often actually reflux or silent reflux.
If your baby's "regression" has been going on for more than 6 weeks with no improvement, it is probably not a regression. It is one of the above.
How long each regression lasts
The general rule is 2 to 6 weeks. The middle weeks are usually the worst. Some averages by age:
- 4 month regression: The disruption typically lasts 2 to 6 weeks, but the underlying sleep change is permanent. Recovery is about building bridging skills.
- 6 month regression: 1 to 3 weeks. Often the shortest.
- 8 to 10 month regression: 2 to 6 weeks. Often the longest and hardest of the first year.
- 12 month regression: 2 to 4 weeks.
- 18 month regression: 2 to 6 weeks. Can feel intense because of the emotional and behavioural component.
- 2 year regression: 2 to 6 weeks. Sometimes longer if triggered by a major life change.
If a regression has continued past 6 weeks with no improvement, revisit the schedule, the routine, and the sleep environment. It has probably become a habit pattern rather than the original regression.
When to see your doctor
Most regressions are self-limiting and require no medical intervention. Some situations warrant a doctor's visit.
- The regression has continued past 6 to 8 weeks with no improvement
- Your baby seems in pain (not just tired or fussy)
- Signs of illness alongside the sleep disruption
- Weight is not being maintained
- Any developmental delays alongside the sleep change
- Persistent breathing issues during sleep (snoring, gasping)
- The disruption is severely affecting family functioning and mental health
Your GP, pediatrician, or health visitor can rule out illness, check developmental progress, and refer to a sleep specialist if needed.
Managing the exhaustion in the meantime
The parent piece is the hardest part of any regression. A few practical points that help:
Split nights with a partner where possible. Even a single 5 to 6 hour consolidated stretch every few nights makes a huge difference.
Do not try to power through alone. Ask for help. Grandparents, friends, hired support. Regressions are a season where outside help pays for itself in your ability to function.
Do not compare yourself to other parents. Most parents omit the regression weeks when they talk about how their baby sleeps. Your baby is not broken.
Trust that it ends. Every regression ends. Usually within a month. Sometimes 6 weeks. It will not be like this forever.
Our post on postpartum exhaustion covers the broader picture on parent wellbeing during hard sleep stages. If you are running on empty, that guide is a better next read than more sleep techniques.
Free download: the Calm Baby Guide
We built a free 14-page guide for parents working through the daily reality of baby sleep at every age, including regressions. Plain language, no fluff. You can download the Calm Baby Guide here, no purchase required.
Frequently asked questions
How long does a sleep regression last?
Most regressions last 2 to 6 weeks. The middle weeks are usually the worst. If a regression has continued past 6 weeks with no improvement, revisit the schedule, environment, and routine, or investigate whether something else is going on.
Does every baby go through every regression?
No. Some babies breeze through several regressions with barely a disruption. Others get hit hard by one specific regression and hardly notice the others. Individual temperament and developmental timing vary hugely.
Should I sleep train during a regression?
Most sleep consultants recommend waiting until the regression has stabilised before starting formal sleep training. The exception is if you are already using gentle sleep support methods, in which case continue what you were doing.
My baby is not at a regression age but sleep is bad. What is going on?
Almost always one of: teething, illness, environment change, schedule problem, or a sleep skill deficit. Regressions are one cause of bad sleep. They are not the only cause. Rule out the alternatives before applying the regression framework.
Can I prevent sleep regressions?
You cannot prevent them, but you can reduce their severity. Consistent daily practice for new physical and cognitive skills, stable wake windows, a consistent bedtime routine, and avoiding new sleep crutches all reduce the intensity of regressions when they arrive.
Is the 4 month regression really permanent?
Yes and no. The underlying brain change (mature sleep architecture) is permanent. What resolves is your baby's ability to bridge the sleep cycle transitions independently. Most babies figure out how to bridge cycles within 2 to 6 weeks of the initial disruption.
Should I feed more during a regression?
Only if your baby is genuinely hungry more often, not as a settling technique. Adding night feeds you did not have before, purely to settle a regression waking, usually creates a new habit that persists past the regression itself.
My baby's sleep is worse than before the regression started. Why?
Because you are still in the regression. Sleep during a regression is often worse than the baseline before it started. This is normal and expected. As the regression clears, sleep returns to the pre-regression pattern, sometimes slightly better because your baby has learned new coping skills through the experience.
Are night wakings during a regression different from hungry wakings?
Yes. A hungry waking rises quickly, is insistent, and does not settle without a feed. A regression waking often has your baby fully alert, sometimes standing or moving in the cot, sometimes just fussing without an obvious specific need. Learning to tell the difference helps you respond appropriately.
The bottom line
Sleep regressions are a normal, predictable part of infant and toddler development. They happen at specific ages because your baby's brain is doing something significant during those windows, and sleep pays the price temporarily.
The universal playbook works for every regression: hold the schedule, hold the routine, do not introduce new sleep crutches, support the new skill during the day, adjust wake windows for the current stage, and use environmental sleep tools that support without creating dependencies.
Most regressions clear in 2 to 6 weeks. The 4 month one is technically permanent (in the underlying brain change) but the sleep disruption resolves within 2 to 6 weeks as your baby learns to bridge sleep cycles. The 8 to 10 month one is often the hardest of the first year because so many developmental things happen at once.
The best mindset for surviving regressions is understanding that they are a sign of your baby's development, not a failure of your parenting. They end. They always end. And with each regression, your baby comes out slightly more capable than before.
Livvewell exists specifically for the daily reality of hard sleep seasons. The tools were designed for exactly this: the settling input at 3am when nothing else is working, the environmental cues that let a fragile sleep hold together, the small windows of hands-free time that let you recover. Regressions are hard. You are not alone in them.