Sleep Training Methods Explained: The Full Spectrum

Sleep training is one of the most searched — and most misunderstood — phrases in baby sleep. Ask ten parents what it means and you'll get ten different answers, most of them coloured by strong emotion. So let's start with what it actually is.

Sleep training is a structured approach to helping a baby learn to fall asleep independently — without being rocked, fed, or held to sleep — and to return to sleep independently after the normal partial awakenings that happen between sleep cycles throughout the night.

That's it. It's not one specific technique. It's not "leaving your baby to cry." It's a spectrum of approaches, ranging from very gentle (a parent stays in the room the entire time) to more direct (a parent leaves and returns at intervals). Most families who work with a sleep consultant end up somewhere in the middle.

It's also worth being clear about what sleep training is not:

• It's not establishing a bedtime routine — that's good sleep hygiene, not training
• It's not creating a dark, quiet sleep environment — that's optimising conditions
• It's not following age-appropriate wake windows — that's basic scheduling
• It's not neglecting your baby or ignoring their needs

Understanding the full spectrum matters because the conversation online tends to collapse everything into two extremes: "do nothing" or "leave them to cry." The reality is far more nuanced than that.

No formal sleep training before four to six months. This is consistent across the research and is non-negotiable in our guidance.

Before four months, babies have immature sleep architecture — just two sleep states rather than the four stages adults cycle through. Their circadian rhythm hasn't established itself yet (melatonin rhythm begins around nine weeks, but mature sleep cycling doesn't start until approximately four months). They also have genuine physiological feeding needs at night — their stomachs are small and they need frequent calories.

The landmark Mindell et al. (2006) systematic review noted that extinction-based techniques are not intended for babies younger than six months. The Gradisar et al. (2016) randomised controlled trial studied babies aged 6-16 months. There is very limited data on formal sleep training under five months.

What is appropriate before four to six months:

• Establishing a consistent bedtime routine
• Practising drowsy-but-awake placement (without pressure if it doesn't work)
• Optimising the sleep environment — dark room, comfortable temperature (16-20 degrees Celsius per Lullaby Trust guidance), white noise
• Learning your baby's sleepy cues
• Following age-appropriate wake windows

These are gentle foundations, not training. They support your baby's developing sleep biology without pushing them beyond what their brain is ready for.

At the gentler end of the spectrum, the parent maintains a high level of presence throughout the process. These methods tend to take longer — typically one to three weeks — but involve less parental distress because you are right there with your baby.

Pick Up/Put Down

Put baby down awake. When they cry, pick them up, soothe them until calm (but not fully asleep), then put them back down. Repeat until they fall asleep. The parent is right there, picking up every time baby cries. Baby will still cry, but they are comforted each time.

This method tends to work well for younger babies (four to six months) and for parents who cannot tolerate any unattended crying. The limitation is that the constant picking up and putting down can become overstimulating for some babies — the repeated motion may actually increase arousal rather than promote sleep.

Gradual Withdrawal (also called the Chair Method or "Camping Out")

The parent sits on a chair beside the cot while baby falls asleep. Every two to three nights, the chair moves further from the cot until eventually it's outside the room. Parental presence is high initially, then gradually reduced over one to three weeks.

This approach is included in the Mindell et al. (2006) systematic review as an effective behavioural intervention, and several NHS trusts across the UK recommend "camping out" as a first-line approach. It works particularly well for babies with separation anxiety (eight months and older) and for parents who want to be present but slowly reduce their involvement.

The trade-off: some babies find a parent's presence without being held frustrating, and it requires real consistency and patience over several weeks.

This is where most families end up — and where the strongest research evidence sits.

Controlled Comforting (also called Timed Checks or Graduated Extinction)

Put baby down awake, leave the room. Return at set intervals — for example, two minutes, then four minutes, then six minutes — to briefly reassure baby with a pat, a shush, or verbal comfort, without picking them up. The intervals may stay the same or increase gradually.

Baby will cry between checks. This is the part that most parents find hardest. But the parent returns regularly — baby is never "forgotten." The process typically takes three to seven nights.

"Controlled crying" is the term most UK parents recognise. "Graduated extinction" is the clinical term used in research. They are essentially the same approach.

The Ferber Method

Similar to controlled comforting, but with a specific, progressively increasing interval schedule — for example, Night 1: three minutes, five minutes, ten minutes; Night 2: five minutes, ten minutes, twelve minutes. Developed by Dr Richard Ferber at Harvard/Boston Children's Hospital.

The Ferber method is a specific version of graduated extinction. It's often conflated with "cry it out" in online discussions, but they are not the same thing. Ferber includes regular check-ins.

What the research says about these methods

The Gradisar et al. (2016) randomised controlled trial — the gold standard study on this topic — tested graduated extinction on babies aged 6-16 months. The results were significant:

• Salivary cortisol (the stress hormone) showed small-to-moderate decreases in the intervention group compared to controls — meaning babies were less stressed, not more
• At the 12-month follow-up, there were no differences in emotional or behavioural problems and no differences in secure versus insecure attachment between groups
• Maternal stress also decreased

The Price et al. (2012) five-year follow-up studied 326 children and found no difference between the sleep training group and the control group across any of 20 measured outcomes — including emotional wellbeing, behaviour, child-parent relationship quality, and stress regulation.

Bedtime Fading

This one sits slightly outside the usual spectrum because it doesn't directly address how baby falls asleep — instead, it addresses when. You temporarily push bedtime later to match when baby naturally falls asleep (reducing the battle), then gradually bring it earlier by fifteen minutes every two to three nights.

Crying is minimal because baby is tired enough to fall asleep quickly. The Gradisar (2016) trial tested bedtime fading alongside graduated extinction — both were effective, and neither showed adverse cortisol or attachment effects. The limitation is that it doesn't directly teach self-settling and may need to be combined with another approach.

Extinction (sometimes called Cry It Out)

Put baby down awake, leave the room, do not return until a predetermined time — usually morning, unless baby is unwell or genuinely hungry. Maximum crying is expected, especially on the first night. The process typically resolves in two to four nights.

This is the method that dominates the online debate — but it's actually the one we least commonly recommend as a starting point. The evidence says it is safe: Mindell et al. (2006) reviewed 52 studies and found 94% reported behavioural interventions were effective, with extinction among them. Price et al. (2012) found no adverse long-term effects at five years.

Our position: we do not recommend extinction as a first-line approach. We acknowledge the evidence says it is safe, but we believe unnecessary distress can be minimised. If a family chooses this method after informed discussion, we support them without judgement — but there are gentler options to try first.

Every method involves some crying — babies cry when things change. The difference is not whether crying happens, but how the parent responds to it and how much presence they maintain.

• Pick up/put down: Maximum parental presence. Baby comforted each time. Slowest (one to three weeks). Best for younger babies and parents who want minimal separation.
• Gradual withdrawal: High presence initially, gradually reduced. Moderate timeline (two to three weeks). Recommended by multiple NHS trusts.
• Controlled comforting / Ferber: Intermittent presence with regular check-ins. Faster (three to seven nights). Strongest research evidence base.
• Bedtime fading: Full parental presence. Minimal crying. Moderate timeline (two to four weeks). Addresses timing rather than settling method.
• Extinction: No parental presence after placement. Fastest (two to four nights). Most parental distress. Least commonly recommended as a first approach.

A few things worth knowing about all of these methods:

• There is no "window" for sleep training that closes. Children can learn independent sleep skills at any age. It's never "too late."
• Many families use a blended approach — starting with gradual withdrawal and moving to timed checks if progress stalls. Flexibility is fine.
• Choosing not to formally sleep train is also a valid choice. Not every family needs to, and not every baby requires it.

The online conversation around sleep training is dominated by two extreme positions, and neither is helpful.

On one side: "Any crying causes brain damage." The studies cited to support this claim — usually the Middlemiss (2012) cortisol study — are about chronic, severe neglect in institutional settings (such as Romanian orphanages), not a responsive, time-limited sleep training process over a few nights. The Gradisar (2016) trial directly measured cortisol and found it decreased.

On the other side: "Crying is fine, babies need to learn." While the evidence supports the safety of sleep training, this dismisses the genuine distress parents feel. It also ignores the importance of responding to babies who are unwell, hungry, or frightened. "Crying is not harmful" does not mean "crying doesn't matter."

The reality is that most families find their way to the middle. They accept that some protest is inevitable during a change, they maintain responsive check-ins, and they adjust based on how their baby responds. The process takes a few days to a couple of weeks for most families.

If you'd like to understand more about the safety evidence specifically, we've covered that in depth in our post on whether sleep training is safe.

If you've read this far, you'll know there is no single "best" method. The right approach depends on your baby's age and temperament, your family's values, how much crying you're comfortable with, and what's practically manageable for your household.

What we can say with confidence is this: sleep training, when done at an appropriate age and with a responsive approach, is supported by evidence and does not harm your baby's development or your attachment relationship. Equally, choosing not to sleep train is a completely valid decision — and one that deserves just as much respect.

The general principles are well understood. But knowing which method suits your baby requires understanding their specific temperament, their current sleep associations, and your family's circumstances. That's where the general information ends and the personalised work begins.