Sleep, by age

Sleep Regression at 18 Months

The age where the bedtime objection comes with a vocabulary — what is genuinely developing at eighteen months, and the nap that has quietly slid an hour later than it started.

9 min read Last checked 10 August 2026 5 sources, all linked

Eighteen months is the one where the child can argue. Language is arriving fast, she has understood that bedtime means you leave, and she now has the words, the legs and the stamina to say so. Underneath the negotiating there is usually a mechanical problem as well: one nap that has drifted too late in the day. Both halves have a boring answer, and neither of them is a discipline problem.

The regression timeline puts this age beside the five others and shows what is developing at each. This page is eighteen months on its own: what has changed since the birthday, why bedtime has turned into a committee meeting, and what to move first.

What this page will not tell you

There is no published eighteen-month event. The NHS, the American Academy of Pediatrics, the NIH and the papers on PubMed do not use the word regression, do not print the list of ages and do not attach a length to any of them — so nobody can honestly tell you how many weeks this runs, and the sites that do have copied a number from each other.

What is published is the useful part: roughly when words, walking and pretending arrive, and that separation anxiety disrupts nights from the second half of the first year until it fades around the second birthday. These are phases in a child rather than clinical events, and the famous list of months is a folk map of when nights commonly turn hard. A map is worth carrying. It is not a measurement, and a hard fortnight at sixteen months or twenty is not less real for missing the map.

The language, and what it does to bedtime

Between eighteen months and two years toddlers learn roughly one new word a week, and the American Academy of Pediatrics prints that as a guideline with wide variation rather than a schedule. Understanding runs well ahead of speaking. A child who has just found words uses them at bedtime, at length, and she now follows what you are doing: coat on, lights down, door.

So the objection is not naughtiness and it is not a habit you have accidentally taught. Separation anxiety is still live at this age — the AAP has it lasting until somewhere around the second birthday — and a toddler who understands that going to bed means you go away has correctly read the situation and does not care for it. That is worth knowing because it decides what actually helps: reassurance and a boundary that stays where it is, rather than a different method.

The shape of the day at eighteen months

One nap, awake stretches of about four to five and a half hours, and 11 to 14 hours of sleep per 24 hours including that nap — that last figure is the sleep consensus the AAP endorses for one- and two-year-olds, and the awake bands are compiled from it rather than quoted from a chart. The wake windows by age page shows the whole arithmetic once; the toddler wake windows guide is the detail for this stretch, and the wake window calculator reads the same bands off an age.

Night waking has not finished, and there is one study large enough to say so honestly. Following more than 1,200 infants to three years, it found two lasting patterns rather than one: about two thirds woke their mothers roughly one night a week throughout, while about a third woke every night at six months and were still waking two nights a week at fifteen months. Neither group is the broken one, and a toddler in the second is not behind.

The single nap that slid too late

This is the most common mechanical cause of an eighteen-month bedtime battle and it is almost never the first thing anybody suspects. The one nap started at half past twelve when she dropped the morning one. Since then it has crept to one, then half past one, and now she wakes at three. Four hours later it is bedtime, she has not been awake long enough to be tired, and you spend an hour and a half on stories, water and the toilet.

The fix is dull and it works: move the nap earlier by about fifteen minutes at a time, hold each move for three or four days, and stop when bedtime stops being a fight. Somewhere around noon to half past twelve suits most toddlers at this age, which is a rule of thumb rather than a finding — nobody publishes a nap timetable any more than they publish a wake-window table. If the nap is also running long — over two hours, with a bedtime past eight that nobody in the house wanted — capping it is a decision about your household rather than a rule about children, and it moves by fifteen minutes too. If she has only recently gone from two naps to one, the two to one transition guide covers the settling-in period, which is bumpier than the change itself.

A fortnight of written-down nap times and bedtimes will show you the drift faster than remembering will — a printed day sheet on the fridge is enough, and it is the difference between guessing and knowing.

The stalling, and the boundary

The third drink, the fourth story, one more song, the sudden urgent thought about the dog. This is the age of the first real testing, and the answer to it is not severity, it is sameness. Pick a routine short enough that you will do it identically when you are exhausted — bath, two books, one song, light off — and then let it be the same every night, in the same order, done by the same steps. Say what comes next before it comes: one more book, then the light goes off.

Answer the returns the same way every time, briefly and warmly, and walk her back. A boundary that moves on the fourth attempt teaches that four attempts is the price; a boundary that never moves is uninteresting within a week or two. None of that means leaving a frightened child alone. It means the answer is the same at the fifth ask as the first.

Molars get blamed for everything

Somebody will tell you it is the molars. It might be. What is worth knowing is that nothing this site is allowed to cite attaches a sleep claim to teething at all — no publisher on the list has measured teething against night waking, so anyone telling you a fortnight of broken nights is definitely teeth is guessing, and so is anyone telling you it is definitely not.

The practical version: a child who is plainly uncomfortable — drooling, chewing, a sore mouth, off her food — can be offered something cold and firm to chew during the day and more comfort at night, and if you are considering pain relief that is a question for your pediatrician, health visitor or pharmacist rather than for a web page. A fever, a rash or a child who seems unwell is not teething until somebody has looked at her.

Cot to bed, and climbing out

Later rather than earlier, unless she is climbing out. A cot is not a punishment and a big bed removes the one thing that was holding the night together; plenty of children stay in a cot well past two and are happier for it. If she is climbing out, though, the argument is over — a fall from the rail is worse than an early move.

If she is climbing and staying in a cot for now, the room becomes the cot. Nothing in it to climb on, nothing that can be pulled over on top of her: furniture strapped to the wall, blind cords tied high and out of reach, the cot away from shelves, windowsills and radiators, and a stairgate at the door or the top of the stairs so a toddler who does get out is contained. That is the same job the cot rails were doing, moved outwards.

A safe room, and when to ring somebody

The safe-sleep guidance is written for the first year and it is worth restating whether or not you still have a baby in the house: babies go down on their back, on a firm flat surface, in a cot clear of blankets, pillows, bumpers and soft toys, and nothing weighted goes on or over a sleeping baby. If there is a younger sibling in the room, that is the rule for her, whatever the toddler in the next bed has.

And if what is happening is not really about bedtime — a child who is unwell, snoring heavily or struggling to breathe in her sleep, losing skills she had, or something that simply frightens you — that is a call to your pediatrician or health visitor rather than a phase to sit out. Nothing here is medical advice and no page can see your child. If you are heading for the next age already, the two year guide covers what changes when the fears arrive, and the twelve month guide is the one behind you.

Questions

Questions parents ask

Not as anything a health body publishes. None of the publishers this site may cite uses the word, prints the list of ages or gives a length for one. What is published is that toddlers learn roughly one new word a week between eighteen months and two years, that separation anxiety runs until around the second birthday, and that night waking is ordinary at this age. Plenty of families do meet a hard few weeks around now, and it is real. It is a stage in a child rather than an event that happens to her.
Nobody has measured it, because nobody who publishes recognises the category. What parents commonly report is two to six weeks, and that figure is unsourced wherever you read it, this page included. The parts that can be timed run longer and more gently: language accelerates across months, and separation anxiety fades somewhere around the second birthday. If a bad stretch has run past a few weeks, stop waiting and look at the day — most often the nap has drifted late, or bedtime has.
Two reasons usually stack. She now understands that bedtime means you leave, and she has the words to object, which is development rather than defiance. And the single nap has probably crept later than it was when she dropped to one, so by bedtime she has not been awake long enough to be tired. Move the nap earlier by about fifteen minutes at a time, hold each move for three or four days, and keep the bedtime routine short and identical. Sameness does more here than firmness does.
Later rather than earlier, unless she is climbing out of the cot. A cot holds the night together and a bed hands an eighteen-month-old the freedom to test every boundary you have; many children stay in one well past two without any harm. If she is climbing, the argument is settled, because a fall from the rail is worse than an early move. Whichever you choose, do the room: furniture strapped to the wall, blind cords high and out of reach, the cot away from shelves and windowsills, and a stairgate at the door or the stairs.
Possibly, and nobody can tell you for certain. No publisher this site is allowed to stand on has measured teething against night waking, so a page that tells you a bad fortnight is definitely the molars is guessing, and so is one that rules it out. What you can act on is comfort: something cold and firm to chew in the day, more settling at night, and a question to your pediatrician, health visitor or pharmacist before giving any pain relief. A fever, a rash or a child who seems unwell is not teething until somebody has looked at her.

Where this comes from

  1. American Academy of Pediatrics (HealthyChildren.org) (2012). Developmental Milestones During the First 2 Years. https://www.healthychildren.org/English/ages-stages/toddler/Pages/Milestones-During-The-First-2-Years.aspx
  2. American Academy of Pediatrics (HealthyChildren.org) (2022). Separation Anxiety and Sleeping. https://www.healthychildren.org/English/healthy-living/sleep/Pages/separation-anxiety-and-sleeping.aspx
  3. Journal of Clinical Sleep Medicine (PubMed Central) (2016). Recommended Amount of Sleep for Pediatric Populations: A Consensus Statement of the American Academy of Sleep Medicine. https://pmc.ncbi.nlm.nih.gov/articles/PMC4877308/
  4. Developmental Psychology (PubMed) (2012). Patterns of developmental change in infants' nighttime sleep awakenings from 6 through 36 months of age. https://pubmed.ncbi.nlm.nih.gov/22448981/
  5. American Academy of Pediatrics (HealthyChildren.org) (2022). How to Keep Your Sleeping Baby Safe: AAP Policy Explained. https://www.healthychildren.org/English/ages-stages/baby/sleep/Pages/A-Parents-Guide-to-Safe-Sleep.aspx

Every link above was opened and read when this page was last updated. Snuggle is not affiliated with any of these organisations and none of them has reviewed this page. Nothing here is medical advice.

The app this site is for

Snuggle does this with your baby’s own pattern.

This page works from an age band, because an age band is all a web page knows about you. Snuggle starts from the same bands and then blends them with what your baby actually did today — how long she lasted, when she went down, how the last nap ran — and shows on the face of the ring how sure it is. Early estimates say they are early.

It logs sleep, feeds, nappies and growth in one tap on a dim red screen, and every caregiver in the household is on the same timeline in their own account. The arithmetic runs on the phone, on your baby’s own history. There are no AI features in it.

In development for iPhone and Android. Not released yet — when it is, it will be on this site.

What Snuggle is