Cry It Out: What Full Extinction Involves and What the Research Shows

Full extinction is the method that gets called cruel by people who have never read the research and easy by people who have never done it at 2 am. It is neither. Here is what it actually involves and what we know about it.

What full extinction is

You put your baby down awake at bedtime. You leave the room. You do not return until a pre-set time (usually morning, or a scheduled night feed if the baby still needs one). If the baby cries, you do not go back in.

That is the method. It is simple to describe and difficult to do. The simplicity is the point: there is no ambiguity for the baby about what happens after you leave. There are no intermittent appearances that restart the arousal cycle. The baby learns, through a few hard nights, that the crib is where sleep happens, and that they can get there on their own.

How it differs from Ferber

In graduated extinction (Ferber), you return at increasing intervals to offer brief reassurance. In full extinction, you do not return. The research shows that both methods produce similar sleep outcomes [1]. Full extinction typically works faster (2-4 nights versus 3-7) because there is no intermittent reinforcement from check-ins.

For some babies, the check-ins in Ferber are calming. For others, the check-ins are stimulating: seeing a parent and then having them leave again escalates the crying rather than reducing it. If your baby seems to get more upset with Ferber check-ins, full extinction may actually involve less total crying [4].

The decision between the two methods is more about parental tolerance than baby outcomes. If you need to check to reassure yourself the baby is safe, Ferber gives you that. If checking makes the situation worse for everyone, full extinction removes the variable.

What the evidence shows

Full extinction has been studied since the early 1990s and is part of every major meta-analysis on infant sleep interventions. The evidence is more extensive than for most parenting decisions that generate less controversy.

Effectiveness: The Mindell et al. 2006 meta-analysis reviewed 52 studies and rated extinction as having “strong” evidence for reducing bedtime resistance and night wakings [1]. Most babies show significant improvement within 2-4 nights. The meta-analysis found that 94% of the reviewed studies reported clinically significant improvements in sleep with behavioral interventions, and full extinction was among the most effective.

The extinction burst: On the first or second night, crying often increases before it decreases. This is a well-documented behavioral phenomenon called an extinction burst: the baby is testing whether the previous response pattern (crying produces a parent) still works. When the expected response does not arrive, the behavior temporarily intensifies before it extinguishes. France and Hudson described this process in infant sleep training specifically, noting that the burst typically peaks on night 2 and drops sharply on night 3 [4].

Understanding the extinction burst is important because it is the point at which most families give up. If you go in on night 2 after 50 minutes of crying, you have taught the baby that 50 minutes of crying works. The next attempt will likely start at that threshold. This is not a reason to use the method or not use it; it is a reason to decide before you start whether you can sustain it.

Long-term outcomes: The Price et al. 5-year follow-up is the most comprehensive long-term study. The researchers tracked 225 families from the original Hiscock trial and assessed children at age 6 on emotional health, behavior, sleep quality, cortisol levels, and parent-child attachment. They found no differences between children who had undergone behavioral sleep training and those who had not, on any measure [2].

Cortisol: The concern that full extinction causes damaging stress is addressed directly by the Gradisar et al. trial. The researchers measured salivary cortisol in infants during sleep training and at 12-month follow-up. There were no significant differences between the graduated extinction group, the bedtime fading group, and the control group [3]. Babies who cried at bedtime did not show elevated cortisol the following day.

Maternal mental health: The Hiscock trial of 328 infants found that sleep interventions improved not only infant sleep but also maternal depression scores. The mothers in the intervention group had significantly lower rates of depression at follow-up [5]. Parental sleep deprivation is not a minor inconvenience; it is a risk factor for postpartum depression, impaired decision-making, and relationship strain.

Who it works for

Full extinction works best for:

• Babies 4-6 months and older whose sleep architecture has matured.
• Parents who can sustain the method consistently. If you go in after 45 minutes of crying, you have taught the baby that crying for 45 minutes works. This is not a judgment; it is an observation about how learning works.
• Families where gentler methods have not worked or where parental sleep deprivation is at a crisis point.
• Babies whose temperament makes check-ins stimulating rather than calming. These are often the babies who get more upset with each Ferber check.

It is less suitable for:

• Parents who cannot tolerate the crying, regardless of what the evidence says. A method you abandon on night 2 at 3 am is worse than a slower method you complete.
• Babies with medical conditions that require nighttime monitoring.
• Situations where shared walls or other children make extended crying impractical.
• Families where one parent supports the method and the other does not. Consistency requires both caregivers to agree.

What to expect night by night

Night 1: Crying at bedtime may last 20-60 minutes. This is the hardest night for parents. Some families find it helpful to leave the house (with a monitor) or to have a partner manage the night. Have a plan for what you will do during the crying: a walk, noise-canceling headphones in another room, a phone call with someone who supports the decision. Going in because you cannot bear it is understandable, but it restarts the process.

Night 2: Often the extinction burst night. Crying may be longer or louder than night 1. This is the night most families abandon the method, which is why knowing it is coming helps. If you made it through night 1, you can make it through night 2. The crying feels like it is escalating because it is: the baby is trying harder before giving up.

Night 3-4: Crying at bedtime typically drops to 5-15 minutes. Middle-of-night wakings begin to resolve. Many families report that the shift feels dramatic: the difference between night 2 and night 3 is often larger than the difference between night 3 and any subsequent night.

Week 2: Most babies are falling asleep within a few minutes of being put down and sleeping longer stretches. Some parents describe a sense of disbelief that it worked, followed by anxiety that it will stop working. It rarely does, provided the routine stays consistent.

Naps and full extinction

Nap training with full extinction is possible but typically harder than bedtime training. The reason is biological: sleep pressure is lower during the day than at night. A baby who falls asleep within minutes at bedtime after extinction training may still cry for 30-40 minutes before a nap.

Many families start with bedtime only and add naps once bedtime is established, usually after 3-5 days. This gives the baby a chance to practice the self-settling skill at the time of day when sleep pressure is highest and the skill is easiest to use.

If the baby has not fallen asleep after 30-45 minutes for a nap attempt, most approaches suggest ending the attempt, getting the baby up for a short period (30-60 minutes), and trying again or adjusting the rest of the day’s schedule. Letting a baby cry for an hour without falling asleep does not teach the skill; it teaches that the crib is an unpleasant place to be.

The limits of the evidence

The research is reassuring, and it is also limited. Most studies follow children for months, not decades. The Price follow-up at age 6 is the longest. The sample sizes are moderate, not enormous. The studies measure what they can measure: cortisol, behavioral outcomes, parent reports, standardized assessments. They cannot measure everything a parent worries about.

What the evidence does show is that the measurable outcomes are good. No study has found harm. Multiple studies have found benefit, for babies and for parents. This is a stronger evidence base than exists for most parenting decisions. It is not a guarantee. No parenting decision comes with one.

A note for tonight

If you are doing this tonight, or thinking about it: the method is hard on parents. The evidence that it does not harm babies is solid. The evidence that sleep deprivation harms parents is also solid [5]. You are allowed to prioritize your family’s functioning, and you are allowed to decide you cannot do this. Both are legitimate outcomes of the same decision.

If you’re concerned about your baby’s sleep or health, speak with your pediatrician or family doctor. The American Academy of Pediatrics has sleep safety guidance at healthychildren.org. If your own sleep deprivation is affecting your daily functioning or mental health, that is also worth raising with your doctor.