Ferber vs. Full Extinction: Which Sleep Training Method Should You Choose

You have decided to sleep train. Now you are staring at two methods that sound almost identical and trying to figure out which one to pick at 9 pm when your baby has been awake for the third time tonight. Here is what actually separates them.

What each method involves

Ferber (graduated extinction): You put the baby down awake and leave. When the baby cries, you return at increasing intervals (3 minutes, then 5, then 10, then 10 again) to offer brief verbal reassurance. You do not pick the baby up. The intervals reset each night. The check-in is short: 30 seconds to a minute. You say something reassuring, pat the mattress or the baby briefly, and leave again.

Full extinction: You put the baby down awake and leave. You do not return until morning (or a scheduled night feed if one is still needed). That is it.

Both methods teach the same skill: falling asleep independently, without a parent present. The difference is whether the parent comes back in during the learning process.

What the evidence says

Both methods receive a “strong” evidence rating in the Mindell et al. 2006 meta-analysis, the most comprehensive review of behavioral sleep interventions, covering 52 studies [1]. The Gradisar 2016 trial found both graduated extinction and bedtime fading effective, with no cortisol differences between groups or from controls [2]. The Price et al. 5-year follow-up tracked 225 families and found no differences at age 6 between children who underwent behavioral sleep training and those who did not, across measures of emotional health, behavior, cortisol, sleep quality, and parent-child attachment [3].

In other words: the research does not show that one method is better or worse for the baby. The differences between them are practical, not clinical.

The real differences

Speed

Full extinction is faster. Most babies show significant improvement within 2-4 nights. Ferber typically takes 3-7 nights [4]. The reason is straightforward: in full extinction, there is no intermittent reinforcement. The baby learns in fewer trials that crying does not produce a parent. In Ferber, the check-ins provide partial reinforcement, which can extend the learning curve.

However, faster is not always better. If the faster method is the one you abandon on night 2, the slower method that you sustain through night 5 produces superior results. Speed only matters if both methods are equally sustainable for your family.

The check-in question

This is where temperament matters, and it is the most useful factor in making the decision. Ferber’s check-ins serve two functions:

1. They reassure the parent that the baby is safe.
2. They are supposed to reassure the baby that the parent still exists.

For some babies, particularly those who calm quickly when a parent appears, the check-ins work as intended. They provide a bridge between needing a parent to fall asleep and falling asleep alone. The baby’s crying decreases after each check because the parent’s brief presence is genuinely soothing.

For other babies, the check-ins are stimulating. Seeing a parent, hearing their voice, and then having them leave again escalates the crying. Each check-in restarts the protest cycle. For these babies, full extinction often involves less total crying than Ferber, which is the opposite of what most parents expect.

You can usually tell which type your baby is within the first night. If crying decreases after the first check-in, even slightly, the checks are working. If crying escalates dramatically after each check, the checks are making things worse.

Parental tolerance

Ferber gives parents something to do. The structured intervals provide a role: you are checking in, following a schedule, actively participating. For many parents, this is psychologically easier than sitting in another room doing nothing while their baby cries. The intervals create a sense of control and a framework for the difficult nights.

Full extinction requires parents to not act, which can be harder than it sounds. There is no schedule, no task, no structured role. The instruction is “do not go in,” and for many parents that feels passive and wrong. The advantage is simplicity: there are no intervals to track, no decisions about when to go in, no moment where you are standing in the hallway debating whether 7 minutes has passed.

The partner question

Both parents need to agree on the method. If one parent can sustain full extinction and the other cannot, the method will be applied inconsistently, which is worse than either method applied consistently. Ferber can be a compromise: the parent who needs to check in can do the checks, while the parent who finds checks stimulating can handle nights when the baby seems to respond better without them.

However, if the parents disagree fundamentally about whether sleep training should happen at all, the method choice is not the actual discussion that needs to happen. That discussion is about whether the family’s current sleep situation is sustainable and what alternatives exist.

How to decide

Choose Ferber if:

• You need a structured role to feel in control of the process.
• Your baby calms when they see you, even briefly.
• You want a slightly more gradual approach than full extinction.
• Both parents need to feel involved in the process.

Choose full extinction if:

• Your baby escalates when you appear and then leave.
• Previous Ferber attempts have not worked because check-ins seemed to make things worse.
• You need the fastest resolution because sleep deprivation is affecting daily functioning.
• You can sustain the method without a structured role.

Consider neither if:

• You cannot sustain either method through the extinction burst (the increase in crying on night 2).
• Your baby is under 4 months old.
• A gentler method (bedtime fading, chair method, gradual withdrawal) would be a better fit for your family. These take longer but involve less crying, and the most effective method is the one you can actually do [5].

What to expect with either method

Night 1: Crying at bedtime, typically 20-60 minutes with either method. This is the assessment night: watch how your baby responds to the first Ferber check-in (if using Ferber) to gauge whether checks are helping or escalating.

Night 2: The extinction burst. Crying may increase. This is the night most parents abandon the method [4]. The increase is a well-documented behavioral phenomenon: the baby is testing whether the old response pattern still works. When the expected response does not arrive, the behavior temporarily intensifies before it extinguishes. Knowing it is coming helps.

Night 3-5: Crying decreases. With full extinction, improvement is often dramatic by night 3. With Ferber, night 4-5. Many families describe the difference between night 2 and night 3 as startling.

Week 2: Most babies are falling asleep with minimal protest regardless of method. Bedtime becomes predictable rather than adversarial.

Common mistakes with both methods

Inconsistency within a night. Going in after 40 minutes of crying on a full extinction night teaches the baby that 40 minutes works. Skipping a Ferber check-in because the baby seems to be settling, then resuming checks when they escalate, sends a mixed signal. Pick the method, follow its rules, and sustain it through the full night.

Changing the method night to night. Ferber on Monday, full extinction on Tuesday, co-sleeping on Wednesday because everyone is exhausted. This is understandable but extends the process. The baby is receiving a different signal each night and cannot learn a consistent pattern.

Starting during a regression. If your baby is in the middle of a developmental disruption (8-month separation anxiety, 12-month walking practice), sleep training may work but the timing is harder. Some families prefer to wait for the regression to pass so the training is not competing with the developmental disruption. Others, particularly those in crisis, need to start now. Both choices are defensible.

Not addressing the sleep environment. Dark room, white noise, appropriate temperature, and a consistent pre-sleep routine all support any sleep training method. If the room is light, stimulating, or inconsistent, the method is working against unnecessary obstacles.

Switching methods

If you start with Ferber and the checks are clearly escalating the crying, switching to full extinction is a reasonable decision. It is better to switch methods deliberately, with a clear reason, than to abandon the process entirely.

The reverse switch, from full extinction to Ferber, is less common but also fine. If you started with full extinction and cannot sustain it, adding structured check-ins gives you a role and a framework.

What does not work is switching back and forth between methods on the same night, or alternating methods on different nights. The baby needs a consistent signal about what happens after bedtime.

A note for tonight

The method matters less than consistency. A parent who picks Ferber and sticks with it for five nights will see better results than a parent who tries full extinction for two nights, switches to Ferber for one night, and then gives up. Pick the one you can sustain, and give it time. Both work [5]. The crying will decrease. The nights will improve. And the decision you are making right now, at this hour, when everything feels impossible, is a decision about your family’s functioning, not your baby’s worth. You are a good parent either way.

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.