If you have decided to help your baby learn to fall asleep independently, the next question is how. The methods range from leaving the room entirely to sitting beside the cot for two weeks. Here is how each one works, what the research actually found about whether they are safe, and how to match a method to your child and your family.
"Sleep training" means helping a baby or young child learn to fall asleep, and fall back asleep, without a parent doing it for them. It is usually considered from around 4 to 6 months, once self-settling is developmentally realistic. It is not something to do with a newborn, and it is not one single method.
Before choosing a method, make sure the basics are right, because none of these approaches work against a schedule that is off. Check that the total sleep on offer matches your child's age, that bedtime is at a realistic time for their body clock, that naps are neither too long nor dropped too early, and that there is no untreated medical issue such as reflux, allergy, eczema, ear infections or snoring. Our bedtime calculator gives you the recommended sleep total and bedtime window for your child's age.
You complete the bedtime routine, put your child down awake, say goodnight and leave, and do not return until a set morning time, apart from genuine safety checks and any night feeds your baby still needs. It is the fastest approach, often producing clear improvement within three or four nights, but many parents find the first two nights very hard to sit through. It suits families who can be fully consistent and struggle with checks that seem to wind their baby up further.
You put your child down awake and check on them at set intervals that get longer over the evening and across the week, for example 3, then 5, then 10 minutes. Checks are brief, boring, and mostly verbal or a hand on the back rather than picking up. This is the most studied behavioural method. Most families see meaningful change within one to two weeks. It is a middle path for parents who want to respond but not to feed or rock to sleep.
You sit on a chair next to the cot or bed until your child falls asleep, giving minimal interaction, then over one to two weeks you move the chair progressively further away and finally out of the room. It is gentler and slower, and it suits older babies and toddlers, children with strong separation anxiety, and parents who cannot tolerate leaving the room. The risk is that a parent in the room becomes its own sleep association, so the key is to keep interaction low and to keep moving the chair.
When your baby cries, you pick them up to calm them, then put them back down drowsy but awake, repeating as many times as it takes. It is very hands-on and can be exhausting, and it tends to work best with younger babies around 4 to 8 months. For many toddlers the repeated picking up is too stimulating.
You temporarily set bedtime to the time your child actually falls asleep now, so they go down tired and settle quickly with little crying, then you move bedtime earlier by about 15 minutes every few nights until you reach the target. It is gentle and low-conflict, and it is especially useful when the main problem is long bedtime battles rather than night waking. It is often combined with a consistent pre-bed routine.
| Method | Typical age | Speed | Best when |
|---|---|---|---|
| Unmodified extinction | 6 months and up | Fast (3-5 nights) | You can be fully consistent; checks make things worse |
| Graduated extinction | 5-6 months and up | 1-2 weeks | You want to respond but not feed or rock to sleep |
| Chair method | Older babies, toddlers, preschoolers | 2-3 weeks | Separation anxiety; you cannot leave the room |
| Pick up, put down | 4-8 months | 1-3 weeks, labour-intensive | Younger babies; you want a no-cry-alone option |
| Bedtime fading | Any age | 1-2 weeks | Bedtime resistance is the main problem |
This is the question most parents actually care about. Behavioural sleep interventions, mainly graduated extinction and bedtime fading, have been tested in a number of randomised controlled trials and reviewed in several systematic reviews. The consistent findings are that these methods improve how quickly children fall asleep and reduce night waking, that they are associated with a modest improvement in parental mood and stress, and that follow-up studies, including one that followed children five years later, have not found harm to the parent-child bond or to children's emotional, behavioural or stress outcomes.
The American Academy of Pediatrics supports behavioural strategies for what is formally called behavioural insomnia of childhood. That said, the trials are mostly in babies over 6 months, methods vary between studies, and "sleep training" is a family decision that also depends on your values, your culture, your child's temperament and how you are coping. Choosing a gentler, slower method, or deciding the timing is not right for your family yet, is a legitimate choice.
Responding to a newborn, feeding a baby who still genuinely needs a night feed, comforting a sick child, going to a child who is in pain, or soothing a frightened toddler after a nightmare are not "giving in". Sleep training is about the routine skill of self-settling in a well child, not about ignoring real needs.
See our full citation list on the Sources & Research page. This article was written and reviewed following our editorial standards.
No method works against a bedtime that is too late or too early. Check the right window for your child's age first.
Calculate Bedtime Now →