Sleep training: what different approaches involve and what the evidence says

3-12 months · Sleep · Reviewed 20 June 2026 · All articles

If you are reading this at 3 am with a baby who has woken for the fourth time tonight, you are not alone. Sleep deprivation in the first year of parenthood is one of the most frequently cited sources of exhaustion and stress for new families, and the question of whether or how to help a baby learn to sleep through the night is one that nearly every parent encounters eventually.

The landscape of advice on infant sleep is wide and sometimes contradictory. You may have been told to let your baby cry it out, or been warned that doing so will cause lasting harm. You may have read about gentle methods that promise no tears, or heard from other parents that what worked for their baby failed completely with yours. All of this is understandable, because infant sleep is genuinely complex, and what suits one family may not suit another.

This article tries to cut through the noise. It covers the major approaches to sleep training, what the research actually says about each, and how different families and cultures around the world think about infant sleep. There is no single right answer here. The goal is to give you accurate information so you can make a decision that works for your baby, your values and your wellbeing.

When babies develop the capacity to self-settle

Before considering any sleep training approach, it helps to understand what babies are developmentally capable of at different ages. The term "self-settling" refers to a baby's ability to fall asleep independently at the start of the night and, crucially, to return to sleep after waking without needing a parent to intervene.

In the first weeks and months, this capacity is very limited. Young babies have immature nervous systems, no circadian rhythm to speak of, and a genuine biological need for frequent feeds throughout the night. Waking every two to three hours in the newborn stage is not a problem to be solved; it is normal infant behaviour.

By around three to four months, a significant shift occurs. A hormone called melatonin begins to be produced in a circadian pattern, meaning babies start to develop a proper day-night rhythm. Sleep cycles become more organised, and babies begin to cycle between light and deep sleep in a way that is closer to adult patterns. This is also the age at which some babies start to show an ability to link sleep cycles independently.

By four to six months, many babies have the developmental readiness to begin learning self-settling skills, provided they have sufficient caloric intake during the day to not need night feeds for nutrition. Most paediatricians and sleep specialists suggest this age window as the earliest appropriate time to consider any formal sleep training approach. The NHS advises that babies often begin to sleep for longer stretches from around six months, though individual variation is considerable.

It is worth being clear that developmental readiness does not mean every baby will achieve long overnight sleep at the same time, or that something is wrong if yours does not. Genetics, temperament, feeding method, and many other factors all play a role.

Cry-it-out (extinction): what it involves and what the evidence shows

Cry-it-out, sometimes called extinction or the extinction method, is probably the most debated sleep training approach. In its simplest form, it involves placing a baby in their cot while drowsy but awake at bedtime, leaving the room, and not returning until morning (or until a pre-set feed time for younger babies). The baby may cry when the parent leaves, sometimes for extended periods on the first several nights, but the theory is that without a parental response to the crying, the baby learns to fall asleep independently.

The name itself can feel alarming to many parents, and the idea of not responding to a crying baby conflicts directly with attachment instincts that are both natural and important. It is worth separating the emotional reaction from what the evidence actually shows.

A substantial body of research, including a landmark randomised controlled trial by Price and colleagues published in Pediatrics in 2012 and their follow-up study in 2016, has found that behavioural extinction does not cause lasting psychological harm, does not raise cortisol (the stress hormone) over time, and does not damage the attachment relationship between parent and child. The 2016 follow-up, which assessed children at age five, found no significant differences in emotional outcomes, behaviour, cortisol levels, or attachment security between those who had been sleep-trained using extinction methods and those in a control group.

In the short term, the method tends to produce results relatively quickly, often within three to five nights. Families who have struggled with severe sleep deprivation sometimes report significant improvements in both baby and parental wellbeing. The main downsides are the distress that many parents experience during the process, and the fact that some babies protest loudly and persistently enough that parents find it unworkable in practice. It is also not suitable for babies who have medical conditions affecting sleep, or for very young infants.

The Ferber method (graduated extinction): what it involves and the evidence

Developed by Dr Richard Ferber and described in his book Solve Your Child's Sleep Problems, the Ferber method is often confused with full extinction but is actually a more graduated approach. It is sometimes called "controlled crying" in NHS and UK guidance, though the terminology varies.

The method involves placing a baby in their cot while awake, leaving the room, and then returning to briefly check on and verbally reassure the baby after progressively longer intervals if they are crying. For example, on the first night a parent might check in after three minutes, then five minutes, then ten minutes for subsequent checks. The intervals are gradually extended over subsequent nights. Crucially, the parent does not pick the baby up during checks; they simply offer a few seconds of verbal reassurance and leave again.

The logic is that the brief check-ins provide enough reassurance to prevent extreme distress, while the progressive extension of intervals teaches the baby that parental presence will return, that the situation is safe, and that falling asleep independently is possible. Many parents find this approach more manageable emotionally than full extinction because they are not leaving the baby entirely alone.

The evidence base for graduated extinction is strong. Multiple randomised controlled trials and systematic reviews have found it to be effective at reducing night waking and settling difficulties, with no evidence of harm to infant wellbeing, attachment, or stress physiology. The NHS acknowledges controlled crying as a recognised sleep training technique for babies over six months, advising parents to check on their baby at intervals to offer reassurance. Studies consistently find it to be one of the faster approaches, with most families seeing meaningful improvement within one to two weeks.

As with extinction, this approach is not for all families. Some parents find the crying during check-in intervals distressing, and some babies escalate their protest when a parent appears and then leaves again, making the experience harder rather than easier. Individual temperament makes a genuine difference here.

Gentler approaches: pick-up-put-down, no-cry and fading

For parents who do not wish to leave their baby to cry, or whose babies respond poorly to extinction or graduated methods, a range of gentler approaches are available. These methods generally take longer to show results, but are effective for many families and may feel more aligned with their values or parenting style.

Pick-up-put-down

Popularised by parenting author Tracy Hogg, the pick-up-put-down method involves placing your baby in the cot, and if they cry, picking them up until they are calm, then putting them down again before they fall fully asleep. This is repeated as many times as needed until the baby settles. The idea is that the baby learns to fall asleep in the cot without the need to be held to sleep, while always having their distress met with a physical response.

The method can be very time-consuming, particularly in the early nights. Some babies find the repeated pick-up-put-down cycle more stimulating than settling, which means it works better for some temperaments than others. It tends to work best for babies aged around four to eight months; older babies may become more agitated by the repeated cycle.

No-cry methods

Approaches like those described by Elizabeth Pantley in The No-Cry Sleep Solution focus on very gradual changes to sleep associations. A common technique is the "pantley pull-off," in which a parent gently detaches a feeding or sucking baby before they are fully asleep, giving them a moment to begin falling asleep independently, then offering comfort again if needed. Over many weeks, the baby gradually learns to fall asleep without the feed or sucking association.

No-cry methods are among the slowest to show results, and they require consistency and patience over weeks rather than days. They tend to suit families who are philosophically opposed to any crying, or who have babies who respond very strongly to any more active intervention. The evidence base is less robust than for extinction and graduated methods, largely because these approaches are harder to study in controlled trials, but they are generally considered safe and appropriate for all ages.

Fading (chair method)

Fading methods involve gradually withdrawing parental presence from the sleep environment over time. In one common version, the parent sits in a chair next to the cot until the baby falls asleep, offering reassurance without picking up. Over subsequent nights, the chair is moved progressively further from the cot, toward the door, and eventually out of the room. The baby learns that the parent is nearby but gradually adjusts to being further away at sleep onset.

This method tends to be slower than graduated extinction but faster than no-cry approaches. It suits families who want to maintain physical presence while still working toward independent settling. It can be harder to implement if a baby becomes dependent on the parent being visible and protests when the chair moves.

Cross-cultural perspectives on infant sleep

It is worth stepping back and recognising that the sleep training debate exists largely within a particular cultural context, one where solo infant sleep in a separate room from an early age is considered the norm. In many parts of the world, this is not the default assumption at all.

Co-sleeping, meaning an infant sharing a sleeping surface with a caregiver, is practised widely across East Asia, South Asia, much of Africa, Latin America, and among many communities globally. Research by James McKenna and colleagues at the University of Notre Dame has documented that mother-infant co-sleeping pairs show synchronised sleep architecture, with frequent brief arousals that may serve a protective function. In Japan, for example, the traditional practice of soine (sleeping together) is deeply culturally embedded and associated with closeness and security. In many Nordic countries, infants nap outdoors in prams even in cold weather, a practice that would raise eyebrows in other settings but which local parents and paediatricians regard as beneficial.

None of this means one approach is universally correct. It does mean that the pressure many parents feel to get a baby sleeping through the night in their own room by a certain age is culturally specific rather than biologically necessary.

At the same time, it is important to be clear about safe sleep. The American Academy of Pediatrics (AAP) and NHS both recommend that babies sleep on a firm, flat surface, on their back, in their own sleep space, ideally in the same room as a caregiver for at least the first six months. The AAP advises against bed-sharing as a general practice due to the risk of suffocation and SIDS, though it acknowledges that some families will choose to co-sleep and provides harm reduction guidance for those who do. If you are considering any form of co-sleeping, reading the AAP safe sleep guidance and speaking with your health visitor is strongly recommended.

The broader cultural point stands, however: there is no single global standard for infant sleep, and the right approach for your family depends on your values, your living situation, your baby's temperament, and your own needs as a parent.

Choosing what works for your family

If you are considering sleep training, it may help to think through a few questions before you start.

Age and developmental readiness. Is your baby old enough? Most specialists suggest waiting until at least four to six months. Has your baby's weight gain been healthy and are they getting enough calories during the day to sustain longer periods without feeding overnight?

Your own values and tolerance. How do you feel about your baby crying? There is no right answer here. Some parents find extinction methods produce results quickly and feel comfortable knowing the evidence does not support harm. Others find any amount of crying deeply distressing and prefer to work more slowly with a gentler method. Both responses are valid.

Consistency. Whatever approach you choose, consistency matters enormously. Starting a method and abandoning it repeatedly can be confusing for a baby and may make sleep difficulties worse over time. Choosing something you can commit to for at least a week is more important than choosing the "best" method on paper.

Your own health and capacity. Parental sleep deprivation is a real health concern and matters as much as your baby's sleep. If you are struggling significantly, it is reasonable to factor your own wellbeing into the decision.

Medical considerations. If your baby has reflux, ear infections, feeding difficulties, or other health issues that may be affecting sleep, it is worth addressing those first. Sleep training is unlikely to resolve difficulties that have an underlying medical cause.

Finally, whatever choice you make, it is worth remembering that nearly all babies eventually learn to sleep through the night. Some do it with sleep training at six months; others get there on their own by 18 months or two years. The timing varies, but the destination is almost always the same.

Frequently asked questions

Is sleep training safe for babies?

The current body of evidence, including studies published in Pediatrics and reviews by the American Academy of Pediatrics, has found no evidence that behavioural sleep training methods cause lasting psychological harm, increase infant stress hormones over time, or damage the parent-child attachment relationship. A 2016 randomised controlled trial by Price and colleagues found no differences in emotional or behavioural outcomes, stress hormones, or attachment between sleep-trained and non-sleep-trained children at age five. That said, individual families differ, and what feels right for one family may not suit another. If you have concerns, speaking with your paediatrician or health visitor is the best next step.

What age can I start sleep training?

Most sleep specialists and paediatric bodies suggest waiting until at least four to six months before attempting any form of sleep training. Before this age, babies have limited capacity for self-regulation and most still genuinely need night feeds. The American Academy of Pediatrics and NHS guidance both emphasise that frequent night waking in the early months is normal and expected. Once a baby has reached a consistent developmental readiness, which is often around four to six months but varies between individuals, sleep training approaches are more likely to be effective and appropriate.

Will sleep training cause my baby lasting stress or harm their attachment to me?

Research to date has not found evidence that sleep training causes lasting stress or damages the attachment relationship between parent and child. The largest randomised controlled trial on this question, published in 2016, followed children to age five and found no significant differences in cortisol levels, attachment security, emotional regulation, or behaviour between sleep-trained and control groups. Attachment is built through thousands of everyday moments of warmth, responsiveness and connection, not by a single sleep training period. If you find any method too distressing for you or your baby, stopping and trying a gentler approach is always a valid choice.

What if none of the sleep training methods work for my family?

It is entirely normal for some families to try a method and find it does not suit their baby's temperament or their own values and wellbeing. Babies differ widely in their sleep patterns, and some children naturally take longer to consolidate overnight sleep regardless of the approach used. If you have tried several methods without success, or if sleep deprivation is significantly affecting your or your baby's health, it is worth speaking to your GP, paediatrician, or a certified paediatric sleep consultant. Some families also find that a combination of approaches, or simply continuing to respond to their baby's needs while waiting for developmental maturity, is the right path for them.

Track sleep, feeds and more in one calm place

Cubby helps you log your baby's sleep patterns, feeds and nappies so you can see what is actually happening, and share it with everyone in your care circle.

Start free

Trusted sources