Bedtime resistance and fear of the dark are the two commonest sleep problems of early childhood, and both are developmental rather than behavioural. A three-year-old who will not settle without a parent in the room has not been badly handled; they have reached the age at which imagination outruns the ability to tell the imagined from the real.
That has a practical consequence. Fears of this kind are not argued away, and the attempt usually makes bedtime longer.
A child's night-time fear is not a mistaken belief to be corrected. It is a feeling to be sat with until it passes.
Children vary, and the useful measure is daytime behaviour rather than the clock. A child who wakes without being dragged out of bed, and who is not falling apart by late afternoon, is getting enough.
Counter-intuitively, an overtired child is harder to settle than a rested one, not easier. Pushing bedtime later to make a child sleepier tends to backfire.
A bedtime routine is a signal, not a sedative. What makes it work is that it is the same sequence, in the same order, at roughly the same time, every night — so the sequence itself starts the process of settling before anything is asked of the child.
The components that matter:
Where bedtime has drifted very late, moving it is better done in steps of five or ten minutes a week than in one jump, because a child put to bed an hour early simply lies awake and the earlier bedtime becomes associated with frustration.
This is the mechanism behind a large share of night waking in children who are otherwise sleeping normally.
Everyone surfaces briefly several times a night. An adult registers the room, recognises it and drops back without remembering any of it. A child does the same — but a child who fell asleep with a parent sitting on the bed wakes to find the parent gone, which is a change in the situation, and calls out. The waking was normal; what makes it a problem is that the conditions for falling asleep are no longer there.
So the useful question is not how to stop a child waking, but what the child needs present in order to fall asleep, and whether that thing will still be there at two in the morning. A soft toy will be. A parent will not.
Fear of the dark, of monsters, of being alone and of burglars or intruders typically appears from around two or three and is at its height between three and six. It coincides with the period in which children develop a vivid imagination and have not yet developed the ability to test what it produces — children under five genuinely cannot reliably separate a dream from a memory.
What helps:
Staying in the room until a child is asleep is reasonable occasionally, on a bad night. Done two nights running it tends to become the new requirement for sleep.
A piece of advice in wide circulation suggests making a "monster repellent" — a spray bottle of water and food colouring — and spraying the bedroom before bed. Similar suggestions include glow-in-the-dark stars described to the child as protective, and searching the wardrobe together to prove it is empty.
All of these confirm the premise. A repellent is only needed where there is something to repel; protection implies a threat; and a nightly search establishes that the wardrobe is the kind of place that requires searching. They often buy a quiet night or two and then entrench the fear, and children frequently escalate — the spray stops working, or the search has to be repeated, or it extends to under the bed and behind the curtains.
The alternative is less satisfying and works better: treat the fear as a feeling rather than a factual claim, do not grant it the dignity of an investigation, and be boringly reliable about coming back.
Letting the dog or cat sleep in a frightened child's room is sometimes suggested on the grounds that it comforts the child and gives the animal a purpose. The comfort is real; the rest needs thinking about. An animal in the bedroom matters for a child with asthma or an allergy, it creates another sleep association that has to be maintained, and a pet that moves about or wants to go out in the night wakes the child. It is a reasonable choice for some households and not a general remedy.
The approach that works is dull repetition. Return the child to bed with as little interaction as possible — no conversation, no negotiation, no lights, minimal eye contact — as many times as it takes, and the same way every night. It is exhausting for several nights and then it usually stops, because getting up has stopped producing anything interesting.
Where a child cannot fall asleep without a parent present, gradual retreat is the method used by health visitors and children's sleep services, and it works by moving the parent out of the room so slowly that each stage is barely a change:
The usual advice is to hold each position for four nights before moving on, and not to move back a stage when the child protests, which is the point at which most attempts fail. Worked through properly it takes around a month. A sticker chart or similar helps with children old enough to care about one.
Methods that involve leaving a child to cry without returning are a separate matter and are genuinely contested; the gradual approaches above are what UK services offer first, and they ask more patience rather than more resolve.
Roughly one child in four has at least one nightmare a week, which makes them ordinary rather than a sign of anything. They occur in REM sleep, so they come later in the night or towards morning, the child wakes properly, is frightened, and remembers the dream.
The response is comfort and reassurance, a short stay, and no interrogation. The dream can be talked about or redrawn the next day. Tiredness, worry, illness and frightening material seen during the day all make them more likely, and what counts as frightening varies wildly between children.
A repeated nightmare with the same theme, particularly one that starts after a specific event, is worth raising with a health visitor or school nurse.
These are routinely confused with nightmares and are a different thing entirely. A night terror comes in the first few hours of sleep rather than towards morning; the child sits up or gets out of bed, screams and appears terrified and inconsolable, and is not awake. They do not recognise anyone, cannot be comforted, and remember nothing of it in the morning. Sleepwalking belongs to the same family and behaves the same way.
Two rules cover the night itself: do not wake the child, and keep them safe until the episode ends on its own, which it will. Those, the scheduled-awakening method that breaks a recurring pattern, and the household safety measures for a child who sleepwalks are set out in full under night terrors.
A child who sleeps well at home may not at a grandparent's house or on a first sleepover, and the reason is usually the routine rather than the place. Taking the comfort object, keeping the same sequence of events in the same order, and giving the host the same short goodnight to use all travel better than reassurance does. A child who is genuinely not ready to stay overnight is better collected than persuaded; being brought home once at ten o'clock does far less harm to the next attempt than a miserable night does.
Most of this resolves with consistency and time. The things that merit help rather than patience:
Health visitors, school nurses and GPs are all reasonable starting points, and several areas run a children's sleep service that takes referrals. The Sleep Charity and the national disability charities provide further support, and sleep difficulties are both commoner and harder to shift in children with a long-term condition or a neurodevelopmental difference, which is a reason to ask for specialist help sooner rather than later.
There is no over-the-counter remedy for a child who will not sleep, and sedating antihistamines should not be used for the purpose. Melatonin is a prescription-only medicine in the UK. It is started by specialist teams, not by GPs, for a defined group of children — chiefly those with autism, ADHD, learning disability, cerebral palsy or complex neurodevelopmental conditions — and only after behavioural measures have been tried properly and treatable causes ruled out. It is reviewed at intervals and withdrawn periodically to check whether it is still needed, because the long-term effects in children are not established. For a typically developing child with bedtime resistance, it is not the answer, and material presenting it as a supplement to be bought and given is describing a different country's rules.