Sleep training methods: a calm comparison of the main approaches
Few topics in parenting spark more debate than sleep training. The phrase alone carries strong opinions, and it can feel impossible to cut through the noise when you're already exhausted. This article sets aside the debate and focuses on what sleep training actually means, which approaches exist, and what the research says. Whatever you choose to do, this is your decision and your family. The aim here is simply to give you clear information so you can make it with confidence.
What sleep training means
Sleep training, at its core, means teaching your baby to fall asleep independently at the start of a sleep period, whether at bedtime or for a nap. When babies fall asleep one way (being rocked, fed, or held), they come to rely on that same condition to resettle when they naturally wake between sleep cycles in the night. All babies cycle between lighter and deeper sleep throughout the night, waking briefly every 45 to 90 minutes. Most adults roll over and drift off again without fully waking. Babies who haven't yet learned to resettle independently will call out for you at each of those transitions.
Sleep training changes what happens at the moment of sleep onset. By helping your baby fall asleep without assistance, you're giving them the skill to reconnect sleep cycles on their own. The specific method you use determines how much you intervene during the learning process, and how long that process takes. Most methods involve some degree of crying because crying is how babies communicate frustration with change, not necessarily distress or harm.
When sleep training is appropriate
Before around 4 to 6 months, sleep training is not developmentally appropriate. A newborn's sleep is governed by a different set of rhythms: they have shorter cycles, more fragmented sleep, and a genuine biological need for night feeds. Their nervous system is also not yet mature enough to manage the self-soothing process that sleep training relies on. Attempting to sleep train too early is unlikely to work, and it may mean extended unnecessary crying for a baby who genuinely needs a response.
Most families who choose to sleep train begin somewhere between 5 and 8 months. By this stage, many babies are developmentally ready to go longer stretches without feeding at night, their sleep cycles are maturing, and the window where learning to settle independently is most responsive tends to be open. That said, it's equally valid to start later. Children can be sleep trained well beyond infancy if that's what your family needs.
If your baby was born prematurely, use their corrected age rather than their birth age when thinking about readiness. It's also worth checking in with your health visitor or doctor before starting if your baby has any medical conditions, is still underweight, or if you have any concerns about their feeding.
Why families choose to sleep train
There's no requirement to sleep train, and this is worth saying clearly. Many families never do it and are perfectly fine. For the families who do choose it, the most common reasons are parental exhaustion that has become unsustainable, babies who wake many times a night and can't resettle without help, situations where one parent is returning to work and needs to function, and occasionally medical reasons where a parent or baby has a health condition that makes fragmented sleep particularly costly.
Sometimes sleep training is chosen after months of trying other approaches that haven't helped. Sometimes it's a decision made earlier, as a practical preference. Neither of these is more valid than the other. The decision about whether to sleep train, and when, belongs entirely to you.
A safe sleep environment comes first
Every sleep training method described below assumes your baby is sleeping in a safe environment. This means a firm, flat sleep surface (a cot or Moses basket mattress, not a sofa or adult bed), your baby always placed on their back to sleep, and nothing loose in the cot, so no pillows, duvets, bumpers, or soft toys around the head. Room-sharing (not bed-sharing) is recommended by the NHS and the American Academy of Pediatrics until at least 6 months, ideally a year.
If you're not sure whether your baby's sleep setup is safe, check the NHS guidance at nhs.uk or the AAP's Safe to Sleep campaign before beginning any sleep training approach.
The main sleep training methods
1. Extinction, sometimes called "cry it out"
Extinction is the most straightforward in its instructions, if not in its emotional experience. You go through your normal bedtime routine, put your baby down drowsy but awake, say goodnight, and leave the room. You don't return during the night for settling, only for genuine illness or safety. Your baby learns to fall asleep on their own because the previous response (you coming in) is no longer available.
In terms of speed, extinction tends to produce results within 3 to 5 nights for many babies. Night one is typically the hardest, with significant crying. Night two is often similar or slightly shorter. By nights three and four, most babies show a marked change. The rapid results are because there's no variability in the response: your baby is not getting intermittent reinforcement from you appearing sometimes, which can actually extend the learning process in other methods.
This approach is the hardest emotionally for most parents. Listening to your baby cry and not responding runs counter to every instinct you have, and that is completely understandable. Some families find it manageable because they can see it working quickly. Others find it incompatible with who they are, and that's equally valid. There is no obligation to use this method, and choosing a gentler approach is not a lesser choice.
2. The Ferber method, also called graduated extinction or "check and console"
The Ferber method was developed by Dr Richard Ferber and described in his book "Solve Your Child's Sleep Problems." The approach has been widely studied and is often what people mean when they say Ferber. You put your baby down awake at bedtime and leave. If they cry, you wait a set interval before going in briefly to comfort them (patting, verbal reassurance) without picking them up. You then leave again. The intervals between your visits increase: you might wait 3 minutes, then 5 minutes, then 10 minutes, with the intervals increasing over subsequent nights.
The brief check-ins are not designed to stop the crying or settle your baby fully. They're meant to reassure your baby (and you) that you're present, and then you leave again before your baby falls asleep. If you stay until they're asleep, the check-in defeats its purpose. This distinction is important and can take some discipline in the moment.
Because the Ferber method is the most studied, it's often cited first in research discussions. Price et al.'s 2012 five-year follow-up study (published in Pediatrics) included both extinction and graduated extinction approaches and found no evidence of harm to children's emotional health, behaviour, or the parent-child relationship at follow-up. Most babies show significant improvement within one to two weeks. The timed check-ins mean this method is slightly slower than full extinction but easier for many parents to tolerate emotionally.
3. Pick up put down
Pick up put down, sometimes shortened to PUPD, takes a different stance. When your baby cries, you go in and pick them up. Once they have calmed, you put them back down in their cot while still awake. If they cry again, you pick them up again and repeat. The aim is to gradually move your baby towards settling in the cot rather than in your arms, while staying responsive to their cries throughout.
This method is considerably more labour-intensive than other approaches. Some nights may involve dozens of pick-ups and put-downs, and the process can take longer than graduated methods because the repeated picking up can sometimes stimulate rather than settle younger or more alert babies. It tends to work better for babies on the younger end of the sleep-training range (around 4 to 6 months) and for babies who find brief physical reassurance genuinely calming rather than over-stimulating.
If you choose this method, consistency matters. Going in, picking up, and then also rocking or feeding to sleep defeats the purpose. The goal is to put your baby down while they are calm but still awake, so they're doing the last piece of the settling on their own. PUPD is not widely studied in the same way Ferber is, but for families who want to stay responsive to every cry, it offers a structured alternative to the approaches that involve any sustained crying without a parental response.
4. The chair method, also called the sleep lady shuffle
With the chair method, you stay in the room while your baby learns to settle, but you gradually reduce your proximity over one to two weeks. On the first few nights, you sit in a chair right next to the cot. You can offer reassurance through a calm voice or gentle touch, but you don't pick up or rock your baby to sleep. Over subsequent nights, you move the chair a little further from the cot: first to the middle of the room, then nearer the door, then just outside the door, and eventually you're not there at all.
The main draw of this method is that your presence is maintained throughout. Many parents find it psychologically easier than any method that involves leaving the room, because you are there and your baby can see or hear you. It can also feel more connected, particularly for parents who feel strongly about not leaving their baby to cry in a room alone.
The tradeoff is that it is one of the slower methods, typically taking two to three weeks before independent settling is well established. There's also a risk that your presence becomes the new sleep association: if your baby is falling asleep looking at you in the chair, the gradual removal of the chair still needs to happen. Some babies also find a parent sitting silently nearby more frustrating, not less, and may protest more than they would without you in the room. It varies widely by temperament. This method suits families who need to feel involved in the process and who can commit to several weeks of gradual change.
5. Fading, also called gradual withdrawal
Fading doesn't require putting your baby down awake from night one. Instead, you gradually reduce the intensity of whatever is currently helping your baby fall asleep, over a period of one to two weeks. If you're currently rocking your baby to sleep, you might rock a little less each night, then just hold, then sway slightly, then stand still, then sit, then put down drowsy, then put down more awake. If you're feeding to sleep, you gradually reduce the amount of feeding before sleep onset, removing the nipple or bottle a little earlier each night until your baby is settling without the active feed.
This approach works with the existing sleep association rather than removing it abruptly, which is why it tends to feel gentler. There's usually less concentrated crying than with extinction-based methods because the change is incremental enough that babies can often adapt without a significant protest. It does require consistency and a clear sense of the steps you're taking: vague fading where you simply try to do "a bit less" each night without a structure can drift and take much longer than it needs to.
Fading works well for families who want to avoid leaving the room or any sustained crying, and who have the time and patience for a slower approach. It tends to suit babies who are sensitive to sudden change. The process takes longer than most other methods, and there can be some nights mid-process where your baby picks up on the change and protests more than usual before adjusting. If you're using fading, having a clear plan for each step before you begin makes a meaningful difference to how smoothly it goes.
6. No-cry approaches
No-cry approaches, popularised by Elizabeth Pantley's book "The No-Cry Sleep Solution," focus on improving the conditions around sleep rather than changing what happens at the moment of crying. The idea is that many babies who sleep poorly are simply overtired, under-stimulated, or in a sleep environment that doesn't help them settle. By optimising these factors consistently, sleep can improve without any method that involves deliberate crying.
Practical no-cry strategies include establishing a consistent, calming bedtime routine (bath, feed, song, dim lights, bed), paying close attention to wake windows so your baby is put down neither overtired nor under-tired, ensuring the sleep environment is appropriately dark and has some gentle white noise to mask household sounds, and timing the last feed so it ends before your baby is fully asleep rather than serving as the settling mechanism. Pantley also describes a technique called the "Pantley pull-off," where you gently remove the nipple or dummy just before your baby is deeply asleep and repeat until they learn to settle without it.
No-cry approaches generally take the longest of all methods and require sustained attention to detail over several weeks. They work best for babies with milder sleep challenges, and for parents whose priority is avoiding any concentrated crying above all else. It's worth being clear-eyed that "no-cry" doesn't always mean completely free of crying: babies may still fuss or cry briefly during transitions even with gentle approaches. But the method avoids any deliberate leaving-to-cry strategy, and for many families that distinction matters enormously. If you try a no-cry approach consistently for several weeks and see no improvement, it may be worth reassessing whether a more structured method is needed.
What the research says
The evidence base on sleep training is more reassuring than the online debate might suggest. The two most frequently cited studies are Price et al. (Pediatrics, 2012), a randomised controlled trial with a five-year follow-up, and Middlemiss et al. (Early Human Development, 2012), which looked at cortisol levels in mothers and babies during sleep training. Neither study found evidence that age-appropriate sleep training causes lasting psychological harm.
The Price study found that children who were sleep trained showed no differences in emotional and behavioural outcomes, sleep patterns, or parental mental health at five-year follow-up compared to children who were not sleep trained. The Middlemiss study raised a more nuanced point: it found that after a few days of extinction training, babies may appear calm outwardly while still showing elevated cortisol internally, at a point when their cortisol was out of sync with their mother's. This finding is often cited as evidence of harm, but the researchers themselves were careful to note that the clinical significance was unclear and that the study was small. The American Academy of Pediatrics reviewed the broader evidence and concluded that behavioural sleep training methods are safe and effective when used at the appropriate developmental stage.
What the research cannot tell you is which method is right for your family. Studies measure group averages, not individual experiences. Your baby's temperament, your own mental health, your family's circumstances, and what you can genuinely sustain all matter. No study is going to tell you how many nights of listening to your baby cry you can manage, or whether your baby is one who will settle quickly or take longer. Those judgments are yours to make, and both the decision to sleep train and the decision not to are supported by the evidence.
It's completely fine not to sleep train
Sleep training is a tool, not an obligation. Responsive settling, where you respond to your baby every time they wake and help them back to sleep, is a valid and loving choice. Development naturally brings more consolidated sleep for many children over time, even without formal sleep training. Some families find responsive settling manageable and even meaningful, particularly those with flexible work schedules or support from a co-parent. Some cultures approach infant sleep in ways that don't involve independent settling at all, and children in those settings grow up without apparent difficulty.
If you're not sleep training because it doesn't feel right for you, that is a good enough reason. If you're not sleep training but you're really struggling with exhaustion, it might be worth talking to your health visitor about what support is available, and reconsidering your options with fresh information. Both of those positions are equally legitimate.
If sleep training isn't working
If you've started a sleep training method and it's not progressing as expected after 1 to 2 weeks, a few things are worth reviewing. Wake windows are one of the most common culprits: if your baby is going down either too tired or not tired enough, the method is working against biology. Check your wake windows against age-appropriate guidance and adjust nap timing if needed.
Consider whether your baby might be unwell. Illness, teething, or a developmental leap can genuinely disrupt sleep training progress. If your baby is running a temperature, cutting a tooth, or going through a notable developmental change (pulling to stand, learning to crawl), it may be worth pausing and restarting once the disruption has passed rather than pushing through.
Also review the sleep environment. Is the room dark enough? Is there enough white noise to cover household sounds? Is the room temperature comfortable? Small environmental factors can have a bigger effect on settling than people expect. And finally, check in on feeding: if your baby has recently had a growth spurt, they may have a genuine caloric need that is driving night waking, and the right response there is a feed rather than a settling approach.
If you've addressed all of the above and sleep is still not improving, a conversation with your health visitor is the right next step. They can help rule out underlying causes and may be able to refer you to a specialist sleep clinic if needed.
Track sleep and see the progress
Cubby's sleep log helps you spot patterns, note what you try and track how your baby's nights are changing.
Start free