It usually starts reasonably. One bad night, one nightmare, one illness, and your child ends up in your bed. Then it happens again. Six months later you are sleeping sideways at the edge of a mattress with a small foot in your ribs, and every attempt to send them back turns into a forty-minute negotiation at eleven at night.
This is one of the most common problems in early childhood, and it is also one of the most fixable. Bedtime resistance in three to six year olds is developmentally normal, not a sign that anything has gone wrong. What follows is a three-week plan built on the interventions that have actually been tested, including one method that held up in a randomised controlled trial, plus an honest account of what does not work.
Why does my child suddenly refuse to sleep alone?
Because two normal things are happening at once. Between three and six, children develop a much stronger imagination and a much stronger drive to test limits, and neither is balanced yet by a mature ability to regulate emotion. The prefrontal cortex, which handles planning and impulse control, is still years from finished.
The result is a child who understands the bedtime rule perfectly and still cannot follow it when tired. Add a new sibling, a house move, starting school or a scary story, and separation at night becomes the thing they push back on, because bedtime is the one moment of the day when a child has your undivided attention and can extend it by simply not going to sleep.
None of this means the habit is permanent. It means the fix has to work with a tired brain rather than rely on reasoning with it.
How much sleep should they be getting anyway?
Children aged three to five need 10 to 13 hours per 24 hours, naps included, according to the consensus recommendation of the American Academy of Sleep Medicine. Six to twelve year olds need 9 to 12 hours. Regularly hitting those numbers is associated with better attention, behaviour, learning, memory and emotional regulation.
One frequent trap: a bedtime that is too early. A child who is put to bed at 7pm but not sleepy until 8pm has an hour of lying awake to fill, and filling it by coming to find you is entirely logical. Track when they actually fall asleep for a week before you assume the problem is defiance.
What is the single highest-value change I can make?
A consistent bedtime routine, done every night. This is not folk advice. A study of 10,085 mothers across 14 countries, published in the journal SLEEP, found a dose-dependent relationship: the more nights per week a family ran a routine, the better the child slept.
The effect size is not small. Children with a nightly routine slept, on average, more than an hour longer per night than children who never had one, and they fell asleep faster and woke less, as the AASM summary of the study notes. Dose-dependent is the operative phrase: five nights a week is genuinely better than three.
A good routine is short, fixed in order, and ends in the child's own room. Four steps is plenty. Bath, pyjamas, two books, lights out. We have written a longer walkthrough in our guide to building a bedtime routine that actually works.
Does the bedtime pass method really work?
Yes, and it is one of the few techniques for this exact problem tested in a randomised controlled trial. Nineteen children aged three to six with bedtime resistance were randomly assigned either to monitoring alone or to monitoring plus a bedtime pass. The pass group left their rooms, called out and cried significantly less, and settled faster, with results holding at a three-month follow-up.
The mechanism is elegant. The child gets one card, exchangeable for a single trip out of the room or one parent visit: a drink, a hug, one more question. After the pass is spent, the parent calmly returns them without conversation. The child keeps a real sense of control, so the panic of being trapped disappears, but the number of interruptions is capped at one.
Crucially, the researchers noted it retains the effectiveness of extinction-based approaches without the extinction burst, the horrible escalation phase that makes cry-it-out unbearable for many families. Parents in the trial rated it highly acceptable. The original paper is in the Journal of Pediatric Psychology.
An honest limitation: nineteen children is a small trial. It is the best-tested option for this specific problem, not a guarantee.
What does the three-week plan look like?
Do not attempt everything at once. Sequence it.
- Week one, fix the timing and the routine. Note when your child actually falls asleep. Set bedtime 15 minutes before that, and run the same four-step routine every single night. Change nothing else. You are building predictability before you ask for anything difficult.
- Week two, introduce the pass. Explain it in daylight, not at bedtime. Let your child decorate the card. One pass per night, spendable on one visit or one trip out. When it is gone, you walk them back silently and calmly. Expect two or three hard nights.
- Week three, shorten your presence. If you have been lying down with them, move to sitting on the bed, then to a chair by the door, then to the doorway. Move one step every two or three nights. If a step fails twice, go back one step rather than abandoning the plan.
Two things make this fail: inconsistency between parents, and giving in at 2am after holding the line at 9pm. Children are excellent at identifying which adult and which hour is negotiable. Agree the script with your partner in advance.
Can a comfort object or companion help?
Often, yes, and for a reason worth understanding. A comfort object works because, unlike a parent, it never leaves. Research on transitional objects suggests children who use one settle more easily and adapt better to unfamiliar situations, and the attachment is developmentally normal from around the first year until roughly age seven.
The value is predictability under the child's own control. A parent comes and goes on adult terms. A bear stays. For a child whose fear is essentially about being alone with their own thoughts in the dark, having something in the bed that responds is worth more than another reassuring speech from the hallway.
That is the role Ted, our talking plush companion, plays at bedtime for some families. It tells a calm story, answers the eleventh question of the evening in a level voice, and does not sigh. Parents set the topics, the tone, the language and the limits in the app, and there is no screen and no camera in the bedroom. We are not going to claim a plush toy solves a sleep problem on its own, because it does not. The routine and the boundary do the work. A companion just makes the boundary easier for a small person to accept. Our approach to what is recorded and what is not is set out on our security page.
If the underlying issue is fear of the dark rather than habit, that needs a slightly different approach, which we covered in our calm plan for children scared of the dark.
When should I stop worrying and ask for help?
Most bedtime resistance resolves with consistency. Some things are worth raising with a doctor rather than waiting out: loud snoring or pauses in breathing during sleep, severe daytime sleepiness despite adequate hours in bed, night terrors that happen most nights, or bedtime anxiety that has spread into daytime worry.
Mayo Clinic's guidance on preschool sleep problems is a sensible reference point for what is typical. And a note on co-sleeping itself: if it is working for your family and everyone is sleeping, it is a cultural preference, not a disorder. This article is for families where it has stopped working.
Frequently asked questions
How do I get my child to sleep in their own bed?
Fix bedtime timing first, run an identical short routine every night, then introduce a bedtime pass that allows one trip out of the room. After the pass is used, return the child calmly and without discussion. Expect noticeable improvement within two to three weeks.
Is it bad for a 5-year-old to sleep in the parents' bed?
Not inherently. Co-sleeping is normal in much of the world and carries no developmental harm for a healthy child of that age. It becomes a problem when it disrupts the sleep of parents or child, and the reason to change it is quality of sleep, not principle.
How long does it take to move a child into their own bed?
Most families see real change in two to three weeks of consistent effort, with the hardest resistance in the first three or four nights. Progress is rarely linear, and illness or holidays usually cause a temporary relapse.
Should I lock the door or use cry-it-out with a preschooler?
No. Locking a bedroom door is unsafe and frightening, and full extinction methods are poorly suited to a child old enough to open doors and negotiate. The bedtime pass was designed specifically to get the benefits of those approaches without the distress.
Does a night light or comfort toy help a child sleep alone?
Frequently, yes. A dim, warm night light and a familiar comfort object both reduce the perceived threat of the dark and give the child something predictable to hold onto. Keep the light low and non-blue, and keep screens out of the bedroom entirely.
Sources cited in this article: the American Academy of Sleep Medicine consensus recommendation on paediatric sleep duration, the SLEEP journal study on bedtime routines in 10,085 families, the AASM summary of that study, the randomised controlled trial of the Bedtime Pass Program in the Journal of Pediatric Psychology, and Mayo Clinic guidance on preschool sleep problems.