Most evenings, the problem is not that your child cannot sleep. It is that they will not stay in bed. One more glass of water, one more question, one more trip to the bathroom, and forty minutes have gone. If you have stood in a dark hallway for the fifth time thinking my child won't go to bed, you are describing something sleep researchers have a name for: bedtime resistance.
It is among the most common things parents raise with a paediatrician, and it is also one of the few evening problems with a decent evidence base behind it. This article covers what is actually happening at that hour, what time a child of three to twelve should be asleep, what a routine changes and how quickly, one specific technique that has been through a randomised controlled trial, and the point at which resistance is telling you something other than "I want five more minutes". The longer guide behind this one is our walkthrough on building a bedtime routine.
Why won't my child go to bed at night?
Usually because bedtime is competing with something better, not because your child is not tired. A four year old can be genuinely sleepy and still refuse, because leaving the room where the rest of the family is feels like missing out. In most children this is a behaviour pattern rather than a sleep disorder.
The French paediatric sleep network Réseau Morphée describes the everyday version of this plainly: a child may want to see a parent who got home late, may want to keep playing, or may simply not understand why sleeping is worth doing. Their guidance on bedtime rituals treats regularity as the main lever, because a predictable sequence lets a child anticipate what comes next without anxiety.
There is a second, more mechanical explanation worth ruling out. A child who has not built up enough sleep pressure will fight bed no matter how calm the room is. The Association Française de Pédiatrie Ambulatoire puts daily physical activity for one to four year olds at around three hours, and flags late or long naps as a frequent culprit. If the afternoon was spent indoors and the nap ran to four o'clock, the eight o'clock refusal is arithmetic, not defiance.
What time should a child this age actually be asleep?
The American Academy of Sleep Medicine recommends 10 to 13 hours per 24 hours for children aged 3 to 5, including naps, and 9 to 12 hours for children aged 6 to 12. Those figures are a consensus target rather than a rule, and they say nothing about bedtime on their own. They become useful the moment you work backwards from the time your child has to get up.
The AASM advisory is short and worth reading in full, and the original document lists the outcomes the ranges are associated with: attention, behaviour, learning, memory and emotional regulation. The American Academy of Pediatrics endorsed the same figures, which is why you see them repeated across paediatric guidance.
One caveat that matters in practice. These are hours of sleep, not hours in bed. Most children take 15 to 30 minutes to fall asleep once the light is out, so a bedtime set exactly at the target is a bedtime that misses it every night. If your evenings run late for structural reasons rather than behavioural ones, our piece on whether bedtime is simply at the wrong time goes through the scheduling side.
Does a bedtime routine really change anything?
Yes, and faster than most parents expect. In a study of 405 mothers and their young children, Jodi Mindell and colleagues found that introducing a consistent nightly routine shortened the time children took to fall asleep and reduced the number and duration of night wakings. Maternal mood improved too, which is not a trivial side effect at nine in the evening.
That study, published in Sleep in 2009, covered 206 infants aged 7 to 18 months and 199 toddlers aged 18 to 36 months across two three-week protocols. A later analysis by the same group, also in Sleep in 2015, found a dose-dependent relationship: more nights with a routine meant better sleep outcomes, and the pattern held across culturally different regions rather than being an artefact of one population. Follow-up work asked how quickly the effect appears and found sleep onset latency improving over the first three nights.
Here is the honest limit. Those samples were infants and toddlers, younger than the three to twelve year olds who do most of the stalling. Extending the finding upward is reasonable, and it matches what paediatricians report, but it is an extension rather than a direct result. A routine is a strong bet, not a guarantee.
What is the bedtime pass, and does it work?
The bedtime pass is a card a child can exchange once per night for one parental visit or one excused trip out of the room. After the card is spent, the parent stops responding to further bids for attention. It was tested in a randomised controlled trial and it worked, with one important qualification about sample size.
Brie Moore, Patrick Friman, Alan Fruzzetti and Ken MacAleese ran the trial with nineteen children aged 3 to 6 who were resisting bedtime, randomly assigned to the pass or to a monitoring control group. Children in the pass condition left their rooms, called out and cried significantly less often than controls, and the time it took them to settle each night dropped significantly. The effects were still there at three-month follow-up. The report in the Journal of Pediatric Psychology also measured how acceptable the method felt: parents rated it more acceptable than either ignoring the child or letting them into the parental bed.
| Step | What it looks like in practice |
|---|---|
| Make the pass | One card, decorated by the child. It is theirs, and the decorating matters because it makes the rule feel negotiated rather than imposed. |
| Explain the deal | The pass buys one thing after lights out: one visit from you, or one trip out of the room. Say what it does not buy. |
| Honour it calmly | When the card is handed over, respond without irritation and without extending the visit. The pass is spent. |
| Hold the line after | Later call-outs get no response. This is the part that does the work, and the part that is hardest for the first three nights. |
Nineteen children is a small trial. The result has been replicated in smaller component studies since, but nobody should present a study of that size as settled science, and the method is built on planned non-response, which some families will find does not fit them. It is offered here as one tested option among several, not as the answer.
What do I do when the stalling has already started?
Shorten your response rather than lengthening the argument. The reliable pattern across behavioural sleep research is that attention, including negative attention, maintains the behaviour it follows. A long negotiation at the door teaches that the door is where conversations happen.
Three things help in the moment. Answer once, in one sentence, and name what comes next: "water is finished, next thing is sleep". Keep the lighting and your voice at the level you want the room to settle to, because raising either resets the child's arousal. And settle the water and bathroom questions before lights out rather than during, since a child who has already had both has nothing left to trade.
Worth holding onto: the goal for the first week is not a silent bedtime. It is a shorter one. A stall that drops from forty minutes to fifteen is the change working, even when it still feels like a stall.
When is refusing to go to bed a sign of something more?
When the resistance is fear rather than negotiation, when it appears suddenly in a child who used to settle easily, or when sleep looks disturbed once the child is actually asleep. Those are different problems with different answers, and a behavioural technique aimed at stalling will not touch them.
Fear of the dark is the most common of these and it follows its own arc, which we go through in our calm bedtime plan for ages 3 to 6. Separation anxiety is the second, and paediatricians on the AFPA's parent platform mpedia describe a version of it that peaks around age three. Loud snoring, mouth breathing, pauses in breathing or a child who is exhausted despite long nights are reasons to talk to a doctor rather than to adjust a routine. The AFPA's overview of non-pharmacological management of children's sleep problems sets out where behavioural approaches stop and clinical assessment starts.
What does a calmer evening actually look like?
Predictable, short, and the same order every night. The research does not point at any particular activity as magic; it points at repetition and at the sequence ending in the child's own room. Three or four steps is enough, and shorter routines are easier to hold on the nights when nobody has the energy for a long one.
A fixed chain works: wash, pyjamas, one story, lights. Keeping the last step quiet and in the bedroom matters more than what fills it. Some families use a story, some a song, some a few minutes of talking about the day. Our guide to how long a bedtime story should be by age covers what that step is actually for.
That last step is also where Ted fits, for the families who use one. It is an interactive plush toy that can take the story on a night when the adult has nothing left, with the topics and the tone set by the parent in the app beforehand. It is a tool for holding the routine, not a replacement for the parent in the room, and it will not settle a child who is frightened or unwell. The evidence here is about repetition, and it applies whether the story comes from a book, from you, or from a toy.
Frequently asked questions
My child won't go to bed. How long should a bedtime routine take?
Around 20 to 30 minutes is typical and enough. Research on nightly routines has found benefits from consistency rather than length, and a routine short enough to repeat on a bad night is worth more than an elaborate one you abandon twice a week.
At what age does bedtime resistance usually stop?
It is most common between about 2 and 6 years old and usually eases as children get older, though it can reappear around changes such as starting school or moving house. The randomised trial of the bedtime pass was run with children aged 3 to 6, which is the band where parents most often ask for help.
Is it bad to lie down with my child until they fall asleep?
It is not harmful, but it does tend to become the condition your child needs in order to fall asleep, including after a night waking. French paediatric guidance suggests leaving the room before the child is fully asleep, so that falling asleep alone is the learned pattern.
Does the bedtime pass work for a child who is scared, not stalling?
It is not designed for that. The pass targets attention-seeking after lights out, and a frightened child needs reassurance rather than planned non-response. Fear of the dark, separation anxiety and disturbed breathing during sleep all call for a different approach, and the last two are worth raising with a doctor.
How long before a new bedtime routine shows results?
Research on introducing a nightly routine found the time children took to fall asleep improving over the first three nights, with fuller effects over two to three weeks. In practice, expect the first few nights to be the hardest, because a rule that used to be negotiable is being tested.
