Sleep training

Sleep training means using a consistent behavioral approach to help a baby learn to fall asleep independently and, over time, to resettle without parental help when they surface between sleep cycles. It is one of the most studied topics in pediatric sleep and also one of the most argued about. The arguments typically involve strong feelings about infant distress, attachment, and what kind of parent responds how — but the actual scientific questions are more specific: do behavioral interventions work, do they cause harm, and what does the evidence show about the different methods 1 2.

What it is and what it is not

Sleep training is not a single method. It is a category of behavioral interventions that range from gradual parental withdrawal with minimal crying to full-extinction approaches that involve sustained infant protest. The word "training" is often loaded — some families prefer terms like sleep coaching or sleep settling — but the underlying mechanism in all the main methods is the same: helping the baby learn that sleep does not require a specific external input (feeding, holding, rocking) every time they surface between sleep cycles 1 3.

Sleep training is also not appropriate before a baby has the developmental readiness to manage the process. Most guidance and most research is based on infants of at least four to six months, with six months being the more common lower boundary for the interventions that involve more sustained protest 1 2.

When babies are typically ready

Developmental readiness is not a switch that flips on a specific date. The factors most clinicians consider:

Age: Most behavioral interventions are studied and practiced in infants from four to six months onward. Before this age, night waking is often still driven by genuine hunger and neurological immaturity rather than learned habit alone.
Weight and feeding: Babies should be gaining well and not relying primarily on night feeds for daily caloric intake. If night feeds are still providing a large share of calories, any approach to reducing night waking needs to pair with daytime feeding adjustments first.
Health: Sleep training during or just after illness makes interpretation almost impossible. A sick or teething baby waking more is not a sleep-training failure — it is a normal response to discomfort that needs a different response.
Family readiness: Everyone in the household needs to understand and be willing to follow the plan. A plan that collapses inconsistently on the third night does not produce a clear learning signal for the baby.

The main methods

These are categories of behavioral approach, not brand names or proprietary systems. They differ primarily in how much protest is involved and how quickly parental presence withdraws.

Graduated extinction (check-and-console)

Often associated with the Ferber method, this approach involves placing the baby in the crib awake at bedtime, leaving the room, and returning at progressively lengthening intervals to offer brief verbal reassurance without picking the baby up. The intervals are typically extended each night. The goal is to give the baby an opportunity to settle while signaling that a caregiver will appear 3 4.

Graduated extinction is one of the most studied approaches and shows strong evidence for reducing night waking and time to sleep onset, typically within a few nights to two weeks 3 5.

Full extinction (unmodified)

Sometimes called "cry it out," this approach involves placing the baby in the crib awake and not returning until morning or a pre-agreed time, with brief safety checks possible but no interaction that resets the settling process. Families who can manage this consistently often find that protest reduces faster than with graduated methods because the signal is unambiguous 3 4.

This approach is effective and has been studied in randomized trials. It does not mean abandoning the baby; it means not providing interactive settling during a defined period of intentional behavioral change 1 4.

Fading and parental withdrawal

These approaches reduce parental presence gradually over days or weeks rather than withdrawing it all at once. The chair method involves the parent sitting near the crib and moving further away over a series of nights. Other variations involve reducing the intensity or duration of settling rather than the timing of check-ins. These approaches involve more parental time and typically slower results, but they involve less sustained crying, which many families find more manageable 3.

The tradeoff is consistency: slow-withdrawal methods only work if the withdrawal actually continues. Stalling partway through tends to reinforce the associations the method was designed to reduce 3.

Bedtime fading

This approach works by temporarily adjusting bedtime to match the baby's current sleep pressure — moving it later so the baby is tired enough to fall asleep quickly — then gradually shifting it earlier. It involves less protest than extinction-based methods and can be used on its own or as preparation for other approaches.

A randomized controlled trial comparing graduated extinction, bedtime fading, and a no-intervention control found that both active interventions improved sleep onset and reduced night waking compared to controls, without negative outcomes for infant wellbeing at follow-up 4.

What the research says

The scientific literature on behavioral sleep interventions is fairly consistent in its main findings, though not every question has been resolved.

Effectiveness: Multiple systematic reviews have found that behavioral interventions substantially improve infant sleep onset time, frequency of night waking, and parental sleep. A widely cited systematic review of behavioral treatments in infants and young children found that extinction and graduated extinction are both well-supported by evidence, with clinically meaningful improvement in most cases 3. A later meta-analysis found that behavioral sleep interventions were associated with fewer sleep problems and improved maternal mood 5.

Safety and attachment: The most common concern is whether sleep training — specifically allowing significant infant crying — causes harm to attachment security, the stress response, or long-term development. The evidence on this point is fairly reassuring. A randomized controlled trial that tracked infants across five years after a behavioral sleep intervention found no significant differences in emotional and behavioral outcomes, sleep patterns, stress indicators, or parent-child relationship quality between children who received the intervention and those who did not 6. A separate randomized trial that directly measured infant cortisol and attachment quality found that neither graduated extinction nor bedtime fading produced elevated cortisol or signs of insecure attachment at 12-month follow-up 4.

The cortisol question: One frequently cited observational study found that while behavioral intervention reduced observable infant distress within a few days, infant cortisol levels remained elevated even after crying stopped, and became desynchronized from maternal cortisol 7. This study is often cited as evidence that sleep training causes hidden stress even when infants appear calm. The study had important limitations: it was observational rather than experimental, used a very small sample, and did not measure longer-term outcomes. More rigorous randomized trials that explicitly measured cortisol at follow-up found no evidence of elevated stress from the intervention 4. The cortisol study is worth understanding; it is not the final word.

Spontaneous resolution: Some families choose to wait. Night waking does decrease over time without intervention in many infants, though the timeline is slower and more variable than with behavioral methods 2. Waiting is a reasonable choice when the family is coping adequately. It is less clearly reasonable when parental sleep deprivation has reached the point of affecting safety or health.

Who sleep training is and is not for

Sleep training is most appropriate for healthy infants roughly four to six months and older who are gaining well and feeding adequately during the day, and for families who have identified a specific sleep problem that is affecting household function and cannot be explained by illness, hunger, or a recent disruption.

It is worth pausing when the baby is currently sick, teething significantly, or in the middle of a major transition like starting daycare or a move. It is also worth pausing if night feeds are still providing a genuinely needed share of calories, or if the adults involved are not aligned on the plan — inconsistency between caregivers undermines any method.

Sleep training is not the right first response when a clinician has not yet ruled out medical contributors. Persistent pain, reflux, feeding problems, and breathing difficulties need to be evaluated and addressed directly, not trained around.

Common pitfalls

Starting before readiness: Attempting extinction-based methods in the first three months produces distress and unpredictable results. The developmental groundwork for self-soothing is not yet in place.

Inconsistency between caregivers: Any behavioral method depends on the baby learning what to expect. If the approach changes night to night or person to person, there is no consistent signal.

Applying sleep training rules during genuine illness or discomfort: A baby who is sick, in pain, or going through a developmental leap is producing a different kind of night waking than a baby settling into a learned habit. These are different situations and need different responses.

Stopping on night two or three: Most families who stop report doing so when crying has typically just peaked and is about to decrease. The first two to three nights of any extinction-based method are usually the hardest, and behavior tends to improve more sharply after that.

Expecting a permanent fix: Sleep training reduces the problem; it does not protect against illness, travel, or developmental disruption. A regression after a week away is not evidence the method failed — it means the baby was affected by change and needs a brief, consistent reset.

Questions to ask your clinician

Is my baby developmentally and nutritionally ready to reduce night feeds as part of this?
Should I adjust the daytime feeding schedule before changing the overnight approach?
Given my baby's temperament and our household situation, which method might fit better?
What would be a warning sign that I should stop and reassess rather than continue?

Sleep training ​

What it is and what it is not ​

When babies are typically ready ​

The main methods ​

Graduated extinction (check-and-console) ​

Full extinction (unmodified) ​

Fading and parental withdrawal ​

Bedtime fading ​

What the research says ​

Who sleep training is and is not for ​

Common pitfalls ​

Questions to ask your clinician ​

Related pages ​

References ​