18 Month Sleep Regression: What Is Really Going On at Bedtime

A lot is happening at eighteen months, and very little of it is about sleep

Start with the development, because that is where nearly every explanation of this stage begins, and because it is the part that is actually documented.

The CDC's milestone checklist for 18 months describes a child who walks without holding on, climbs on and off a couch or chair without help, tries to say three or more words besides "mama" or "dada", follows a one-step instruction without gestures, and, in the line most relevant to bedtime, "moves away from you, but looks to make sure you are close by". Read that last one again with the cot rail in mind. A child who has just learned that they can walk away from you, and who still needs to check that you are there, is a child for whom the moment you walk away from them has become interesting.

The American Academy of Pediatrics makes the same point from the other side. Its guidance on separation anxiety explains that separation anxiety develops after a child understands object permanence, that most babies show a more robust form around nine months, and that toddlers start showing challenges with separation at fifteen or eighteen months. It also notes that separations are harder when a child is hungry, tired or sick, which describes bedtime with some accuracy.

Then there is independence. The Sleep Foundation's page on this stage lists bedtime resistance tied to a growing sense of independence or evening overstimulation, restlessness from new mobility, separation anxiety (which it calls a major contributor to sleep problems in toddlers), teething discomfort, and changes to the sleep schedule. Its conclusion is that these disruptions typically do not have one single cause.

Teeth deserve a sentence of their own, because they get blamed for a great deal. The AAP's teething guidance says teething can cause swollen, tender gums and mild pain and can disrupt everyone's rest, but that fever is a false symptom of teething and that teething discomfort is generally not severe enough to make a baby cry more than usual. If a child is feverish or plainly unwell, look for another cause, and do not reach for numbing gels or teething necklaces, both of which that page warns against.

Is this a regression at all?

The word suggests going backwards, and that is the first thing worth questioning. A child who has learned to walk, to call out, and to notice that you have left is not going backwards. Their sleep has hit the same change everything else has.

It also suggests something universal, as if every child arrives at eighteen months and falls off a cliff. The sources do not support that. The Sleep Foundation describes regressions as phases that can occur in any child but are not destined to occur in every child, and states plainly that many toddlers do not have one at this age. If yours is sleeping well, nothing is pending.

And it suggests that good toddler sleep means no waking. That is not what the research shows either. A systematic review of normal sleep patterns by Galland and colleagues, pooling diary and questionnaire data from 34 studies, gave a reference range of zero to 2.5 night wakings per night for children aged one to two years. A large US study that followed over 1,200 children by Weinraub and colleagues found one group whose mothers still reported awakenings on about two nights a week at fifteen months, falling to about one night a week by twenty-four. Some waking in the second year is inside the normal range, not a sign that something has broken.

None of that makes a run of bad nights easier. It changes what you are trying to do about them. You are not fixing a fault. You are steering a child through a stretch where bedtime became more interesting and separation became harder, while keeping the rest of the day steady enough that it does not add to the problem. The sleep regression chart on this site sets the eighteen-month row beside the others, with a column on how well supported each age actually is.

Three shapes the nights usually take

Most families describing a hard stretch at this age are describing one of a small number of patterns, and telling them apart is more useful than giving the whole thing one name. The descriptions below are drawn from the causes the Sleep Foundation and the AAP list, not from any measured frequency.

The first is a bedtime that turns into a negotiation. The routine runs normally, the lights go out, and then come the requests: water, another book, a different toy, you. The child is not distressed so much as busy. This is the independence half of the picture, and it tends to respond best to a routine that is unhurried but not open to renegotiation, the same steps in the same order, ending at the same point every night. A large questionnaire study by Mindell and colleagues across thirteen countries and regions found consistent bedtime routines associated with shorter time to fall asleep, fewer night wakings and longer sleep in young children. It is an association, not a trial, but it points the same way as everything else.

The second is protest at the moment you leave, and at wakings later in the night, where the child settles quickly once you appear. This is the separation half. The AAP's separation guidance is written for daytime drop-offs, but its core advice carries over: keep the goodbye short and sweet, use the same ritual at the same time each day, stick to your promise of return, and practise short separations so that coming back becomes something the child already expects.

The third is a child who is simply wakeful: long stretches of chatting, standing and practising in the cot, early waking, a nap that runs short or not at all. This is often a timing problem as much as a developmental one, and the next section deals with it.

Whichever shape fits, a few things the sources agree on hold across all three:

• Keep the same wind-down routine, in the same order, every night, including weekends
• Respond the same way each time, so the child learns what bedtime now looks like
• Keep daytime separations short and cheerful, so returning is a pattern they already know
• Check the day's sleep total against the published range before changing anything at night
• Hold off on big changes, such as a move to a bed or a new room, until the nights settle
• If waking comes with fever, pain, snoring or laboured breathing, treat it as a medical question

The day behind the night

When the nights go wrong it is natural to look only at the nights. Very often the cause sits earlier.

Start with totals. The American Academy of Sleep Medicine's pediatric consensus, which the AAP endorses, recommends 11 to 14 hours of sleep per 24 hours for children aged one to two years, naps included. The Sleep Foundation describes that total as usually made up of one night-time sleep and one daytime nap at this age. Add up a few typical days. If the total is landing well inside that range, the problem is probably not quantity. If it is landing near or below the bottom, a tired child is a harder child to settle, and the fix may be a bedtime that comes earlier, not later.

Then look at the nap. A child still on two naps whose afternoon nap is pushing bedtime back is a common source of a long, chatty bedtime; a child who has recently dropped to one nap and is running a deficit is a common source of an overtired one. The two look different. The first is cheerful and unbothered at 8pm; the second is fraught and falls apart late in the routine. The mechanics of that change, and the bridge weeks in the middle of it, are on the 2 to 1 nap transition page.

Finally, check the morning. A wake time that floats by an hour from day to day moves everything after it, nap and bedtime included, and makes it nearly impossible to tell whether a bad night was caused by development or by a shifted day. Holding the morning wake within a narrow window for a week is the cheapest experiment available, and it costs nothing in tears.

None of this needs a stopwatch. The homepage has the schedule pages that work through the arithmetic for younger ages, and the same logic of counting from a fixed morning carries straight on into the second year.

How long it lasts, and the point to stop calling it a regression

The Sleep Foundation's answer is that the symptoms of this stage rarely last more than a few weeks, and that sleep problems continuing for a month or more are worth raising with a pediatrician. That is about as precise as any honest source gets. No study has timed an eighteen-month regression, because no study has defined one; the durations repeated online are recollections, not measurements.

A month is a useful line for another reason. Developmental disruption tends to ease as the skill or the separation becomes familiar. A pattern that keeps going, or gets worse, past that point is a sign the explanation was wrong, and it is time to look at other causes rather than keep waiting. Those include an ear infection or other illness, reflux, a schedule that drifted and never came back, and sleep-disordered breathing, which can show up as snoring, mouth-breathing, restless sleep or pauses in breathing. Those are conversations with your pediatrician, not things to sit out.

It is also worth separating this stage from the decision about sleep training, because the two often get tangled at this age. Some families use a stretch like this to change how bedtime works; others wait for it to pass and then decide. If you are weighing a structured approach, the Ferber method page sets out what one widely used option involves and what the research on it does and does not show. There is no need to decide during the worst week.

And keep sleep safety in view while everyone is tired, especially with a child who can now climb. For safe-sleep guidance, the AAP's sleep pages are the source to read, not a summary here. The sleep regressions hub collects the other stages if a later one turns up.