Weissbluth Sleep Training: The Extinction Method Explained

Healthy Sleep Habits, Happy Child: the core philosophy

Marc Weissbluth is a paediatric sleep researcher who spent decades studying infant sleep at the Ann & Robert H. Lurie Children's Hospital of Chicago. His book, first published in 1987 and updated multiple times since, argues that sleep is a biological need as fundamental as nutrition, and that chronic sleep deprivation in infancy has measurable negative consequences for development, mood, and learning.

Weissbluth's central insight is that overtired babies sleep worse, not better. When a baby stays awake past their optimal sleep window, cortisol (the stress hormone) rises, making it harder to fall asleep and stay asleep. The solution is to watch carefully for the first signs of tiredness — eye rubbing, pulling at ears, yawning, loss of interest in toys — and act on them immediately by starting the sleep routine.

Weissbluth is also one of the strongest advocates for the biological necessity of consolidated sleep. He argues that broken sleep, even if it totals an adequate number of hours, does not produce the same restorative benefits as uninterrupted sleep. Teaching a baby to sleep through the night is, in his view, a developmental gift, not an imposition.

What extinction means — and why it works

Weissbluth's recommended primary method is unmodified extinction: you put the baby down awake after a consistent bedtime routine and do not return until morning (or a set feed time if the baby is still young enough to need it). There are no check-ins, no soothing visits, no responses to crying.

This sounds harsh, but there is a clear biological explanation for why it works faster than graduated approaches. When a baby cries at night and a parent consistently responds with rocking, feeding, or any form of soothing, the baby learns — at a very basic neurological level — that crying produces that response. The behaviour is reinforced. Without any response, the reinforcing link is broken and the crying behaviour extinguishes relatively quickly.

For some babies, the Ferber method's check-ins actually extend crying because each parental visit reminds the baby that crying brings parents and raises anticipation. Weissbluth found that for these babies, no check-ins at all produced less total crying time over the week than graduated approaches.

Extinction works because it removes the reinforcer, not because it ignores the baby. The baby is safe, fed, clean, and in a comfortable sleep environment. The crying reflects protest, not genuine need.

When to start: age 4–6 months and the sleep window

Weissbluth distinguishes between different starting points:

• 4 months: This is the earliest he recommends any form of sleep training. By 4 months, most babies have a developing circadian rhythm and can begin to consolidate sleep. Weissbluth often suggests simply establishing a consistent bedtime around 6–8 p.m. and letting the baby fall asleep without active intervention, which for some families is enough.
• 5–6 months: The most common recommended window for full extinction. By now most healthy full-term babies do not have a nutritional need for night feeds, though many still wake from habit. Night sleep consolidation is neurologically possible.
• 6–12 months: Still effective, though the process may take a day or two longer as habits are more ingrained.
• 12 months+: Possible but harder due to increasing object permanence and separation anxiety. Weissbluth recommends not waiting this long if sleep is already a significant problem.

Regardless of age, Weissbluth emphasises that an early bedtime (often 6–8 p.m.) is one of the most powerful tools for improving baby sleep, counter-intuitive as that may feel to parents who worry an early bedtime means an early wake-up.

Nap extinction vs night extinction

Weissbluth addresses naps and night sleep separately. His guidance:

• Night sleep first: Begin extinction at bedtime and for night wakings. Night sleep responds faster to training because of stronger homeostatic sleep pressure and more mature circadian regulation.
• Nap extinction: Apply the same extinction principle at nap time — put down awake, leave the room. However, cap the nap attempt at 60 minutes. If the baby has not fallen asleep, get them up and try again at the next nap window. Do not let a baby lie awake for hours in the hope they will eventually sleep.
• Nap schedule by age: By 3–4 months, most babies have two to three naps per day. By 6–9 months, the third nap is usually dropping. By 15–18 months, most babies transition to one nap. Weissbluth is emphatic that protecting nap opportunities — even if they are not yet sleeping independently — is essential.

A common pattern is for night sleep to improve within one week of extinction while naps take two to four weeks longer to consolidate, because daytime sleep pressure is lower and the biological drive to nap is easier for a baby to override through protest.

Modified extinction: a middle-ground option

Weissbluth acknowledges that not every family can tolerate full extinction and offers a modified version: a single check-in after an agreed period (e.g. 20 minutes) on the first night only, or maintaining one dream feed (a feed given to a sleeping baby at the parent's bedtime) while applying extinction to all other night wakings. He is clear that this is a compromise — it will likely add a day or two to the timeline — but it is preferable to abandoning the process entirely.

For parents who find extinction genuinely unbearable, Weissbluth recommends the Ferber method as a valid alternative, while noting that some babies respond less well to it. He is explicit that the goal — a well-rested baby who can sleep independently — matters more than the specific path taken to get there.

Frequently asked questions

How does Weissbluth sleep training compare to Ferber?

Both approaches teach independent sleep, but they differ in method. The Ferber method uses timed check-ins — a parent returns at increasing intervals to briefly reassure the baby without picking them up. Weissbluth's extinction method involves no check-ins at all after the baby is placed in the cot. Weissbluth argues that for some babies, parental check-ins extend and intensify crying rather than reassuring. Full extinction typically produces results in 3–7 nights; Ferber usually takes 1–2 weeks. Neither has been found to cause lasting developmental harm in research.

Is there an age limit for cry-it-out sleep training?

There is no strict upper age limit, but most experts and Weissbluth himself recommend starting full extinction no later than 12–18 months. After this age, babies develop stronger object permanence and separation anxiety, which can make extinction significantly harder. Starting between 4 and 6 months — when sleep cycles are maturing but object permanence is not yet fully developed — generally produces the fastest results with the least total crying.

How long will my baby cry during Weissbluth extinction?

Crying duration varies widely by baby. The typical pattern: night one may involve 20–60 minutes of crying; night two is often the peak; nights three through five usually show a marked decrease. By night five to seven, most babies fall asleep within 10–15 minutes. A small number of babies take up to two weeks. The absence of check-ins means there are no escalation cycles — for some babies this results in less total crying than Ferber.

Does letting a baby cry-it-out increase SIDS risk?

No. There is no evidence that extinction-based sleep training increases the risk of SIDS. SIDS risk is associated with sleeping environment factors (soft bedding, overheating, smoke exposure, unsafe sleep surfaces) and not with whether a baby cries before falling asleep. The American Academy of Pediatrics safe sleep guidelines focus on the sleep environment, not on the method used to help a baby fall asleep. Follow safe sleep guidelines regardless of your sleep training approach.

Who is Dr. Marc Weissbluth?

Dr. Marc Weissbluth is a pediatric sleep researcher and clinical professor of pediatrics at Northwestern University Feinberg School of Medicine. He has studied infant and child sleep for decades and treated thousands of children with sleep disorders at his clinic in Chicago.

His book Healthy Sleep Habits, Happy Child, first published in 1987 and updated multiple times since, is one of the best-selling parenting books of all time. It is unusual among parenting books for its depth of clinical data: Weissbluth draws on longitudinal studies of his own patient populations as well as broader sleep research to build his recommendations.

Unlike Ferber, who developed a structured interval protocol, Weissbluth's approach is less about a specific schedule and more about three core principles: early bedtimes, age-appropriate nap schedules, and uninterrupted sleep learning. His fundamental argument is that overtiredness is the root cause of most infant sleep problems, and that well-rested babies are easier to sleep train and sleep better overall.

Extinction vs graduated extinction: what's the difference?

The Weissbluth method belongs to the full extinction category of sleep training, as opposed to the graduated extinction (Ferber) approach. Understanding this distinction is important:

• Full extinction (Weissbluth): After completing the bedtime routine and placing the baby in the crib awake, the parent does not return to the room until morning or until a scheduled and necessary feed time. There are no timed check-ins. The baby is given the complete opportunity to learn self-settling without the complicating factor of periodic parental reappearances.
• Graduated extinction (Ferber): The parent returns at progressively increasing timed intervals to briefly reassure the baby. The check-ins do not involve picking up or nursing — they are brief reassurances. The waiting intervals increase across each night and across nights.

The reason Weissbluth advocates for no check-ins is behavioral: in his clinical observation, check-ins can sometimes paradoxically increase crying by giving the baby a signal that crying summons parental presence. The baby learns "if I cry long enough, they appear" — which can motivate more sustained crying rather than less. This is not universal, but it is a documented pattern in some infants.

Research comparing the two approaches (notably the 2016 Hiscock et al. RCT in Pediatrics) found no significant differences in outcomes between full and graduated extinction. Both methods work. The choice between them comes down to parental tolerance and individual infant temperament.

Age guidelines: when to start

Weissbluth discusses sleep habits beginning from birth, but draws careful distinctions between age-appropriate advice for different developmental stages:

• Birth to 4 months: Weissbluth focuses on establishing healthy sleep habits through timing, not training. He recommends soothing to sleep actively during this period, watching for drowsy cues, and protecting nap schedules. No formal extinction training is recommended.
• 4 months: This is the earliest Weissbluth considers beginning extinction training, and only in situations where the baby is gaining weight well, has no medical issues, and the pediatrician has cleared the reduction of night feeds. Many sleep experts consider 4 months the lower boundary and prefer 5–6 months.
• 5–6 months: The most common and widely supported starting point for full extinction. At this age, babies typically can go 5–7 hours without a genuine nutritional feed, have completed the 4-month sleep regression, and have the neurological maturity to self-settle.
• 6 months and older: Full extinction remains effective at any age. Weissbluth's data shows good outcomes even when families begin at 8–12 months, though toddlers have more persistence and the first nights may be longer.

If your baby was premature, use corrected age for developmental readiness. Always consult your pediatrician before reducing night feeds, particularly if your baby has any weight or growth concerns.

What happens during Weissbluth sleep training

The method has four essential elements that all work together. Missing any one of them significantly reduces effectiveness:

1. Early bedtime

Weissbluth's most distinctive recommendation is the early bedtime — earlier than most parents instinctively expect. For babies 4–12 months, he recommends bedtime between 6:00 and 7:30 PM. For many families conditioned to keeping babies up until 9 or 10 PM to "tire them out," this is counterintuitive.

The rationale is cortisol biology. When a baby stays awake past their optimal sleep window, cortisol (the stress hormone) spikes to keep them alert. Cortisol has a half-life — once elevated, it takes hours to clear the system. A baby put down at 9 PM after a cortisol spike has physiologically more difficulty self-settling than the same baby put down at 6:30 PM before the spike occurs. Weissbluth observes that many "difficult" sleep training situations are simply overtired babies being put down too late.

2. Consistent bedtime routine

A 20–30 minute predictable sequence signals to the baby's nervous system that sleep is approaching. A typical Weissbluth-aligned routine: bath → feed (not to full sleep) → book or song → crib. The feed should occur early enough in the routine that the baby does not fall fully asleep while nursing or taking a bottle, maintaining the "drowsy but awake" placement that makes self-settling possible.

3. No check-ins

After placing the baby in the crib awake, the parent leaves and does not return (barring genuine physical emergency) until morning or a pre-decided scheduled feed time. Any check-in, even a brief one, resets the emotional arc the baby is working through. This is the hardest part for most parents and the most important part for the method to work quickly.

Many families find it helpful to have the non-primary caregiver be the one to do the check-in monitoring. If you must check on your baby, use a monitor — don't open the door unless you genuinely suspect something is physically wrong.

4. Protected nap schedule

Weissbluth emphasizes that night sleep and nap sleep are interdependent. A baby with chaotic or insufficient daytime naps will be overtired at bedtime, making extinction training harder. Before beginning, establish the appropriate number and timing of naps for your baby's age and follow them consistently for at least a week.

What to expect night by night

The CIO timeline is typically faster than graduated extinction — most families see substantial change within 3 nights:

These are averages. Some babies take longer — up to 1–2 weeks for consistent settling. Some babies show very little improvement on nights 1–2 and then dramatically improve on night 3–4. The pattern is highly individual.

Research on CIO: what does the science actually say?

The cry it out debate generates strong feelings, often disconnected from what the research actually shows. Here is an honest summary of the key evidence:

Evidence supporting safety

• Price et al. (2012), Pediatrics: A randomized controlled trial comparing behavioral sleep interventions (including graduated extinction) to control groups. No significant differences found in cortisol, child behavior, emotional development, or parent–child attachment at 5-year follow-up. This is one of the most robust studies on the topic.
• Hiscock et al. (2014 & 2016), Pediatrics: Australian RCT comparing graduated extinction, bedtime fading, and control. Found improvements in infant sleep and maternal depression with no adverse child outcomes at 6-year follow-up. Both extinction and fading groups performed better than control on sleep measures.
• Gradisar et al. (2016), Pediatrics: Compared full extinction, graduated extinction, bedtime fading, and control in a randomized trial. All three intervention groups improved sleep without elevated cortisol or attachment insecurity. Parental wellbeing also improved significantly in the intervention groups.

The Middlemiss 2012 cortisol study — what it actually showed

A 2012 study by Middlemiss et al. is frequently cited in anti-CIO discussions. The study found that after several nights of sleep training, babies' cortisol levels remained elevated even after they stopped crying outwardly. Critics interpreted this as evidence of hidden stress after apparent self-settling.

Important limitations of this study: it was small (25 mothers and infants), had no control group, measured cortisol only at one timepoint, and the cortisol levels themselves were not in a pathologically elevated range. The authors themselves noted the findings were preliminary and required replication. No subsequent study has replicated this specific finding in a controlled design. The major RCTs (Price 2012, Gradisar 2016) with larger samples and control groups found no evidence of elevated cortisol.

Weissbluth M., "Healthy Sleep Habits, Happy Child" — source note

Weissbluth draws extensively on his own longitudinal patient data as well as published sleep research. While his book is not a peer-reviewed publication, the clinical observations he documents regarding early bedtimes, sleep deprivation effects, and the relationship between overtiredness and fragmented sleep are consistent with independent research findings.

The emotional impact debate

The most emotional aspect of the CIO debate is not usually about the research — it's about how it feels to listen to your baby cry without intervening. This is a real consideration that deserves honest treatment.

Most parents find nights 1 and 2 extremely difficult. The experience of listening to sustained crying without responding goes against deep parental instincts. Many families report significant parental distress during these nights, even when they understand intellectually that the method is time-limited.

The research consistently shows that parental distress during CIO does not predict worse child outcomes. However, it does predict whether families complete the method consistently — and consistency is the single most important variable for success. A family that abandons CIO after night 1 and restarts a week later experiences more total crying over a longer period than a family that completes 3 consistent nights.

The honest guidance: if you genuinely cannot sustain the experience of full extinction without intervening, choose a different method. The Ferber method or chair method may produce the same long-term outcome with a process that is more tolerable for you specifically. A method you can maintain consistently is always superior to a theoretically faster method you cannot maintain.

Contraindications and practical tips

When not to use the Weissbluth method:

• Babies under 4–5 months (chronological age; corrected age if premature)
• Babies with medical conditions, ongoing illness, reflux requiring treatment, or any condition causing genuine nighttime discomfort
• Babies not gaining weight adequately who still require frequent night feeds
• During periods of major transition: illness, travel, moving house, new caregiver, new sibling within the last 2–4 weeks
• Families in shared housing situations where sustained crying is not feasible

Practical tips for success:

• Pick a stable week. Both parents at home, no travel, no illness expected, no major change imminent. The first week of CIO is the hardest and requires the most consistent response.
• Move bedtime earlier before you start. For 3–5 days before the first training night, shift bedtime 30 minutes earlier. Weissbluth's observation is that a well-rested baby on the first training night settles significantly faster than an overtired one.
• Don't look at the monitor video. Use audio only if possible. Watching your baby cry while knowing you are not going in amplifies parental distress without providing any safety benefit.
• Agree in advance on night feed policy. If a night feed is still needed, schedule it at a fixed clock time rather than in response to crying. This separates nutrition from the sleep association of being fed to sleep.
• Exit the house if necessary. Some parents find that going for a walk, sitting in the car, or visiting a neighbor for 30 minutes on night 1 is what allows them to stay consistent. Do whatever you need to do as long as your baby is physically safe in their crib.

Weissbluth vs other sleep training methods

For a full comparison across all methods, see our complete sleep training methods guide. For the Ferber alternative with timed check-ins, see our Ferber method guide.

Related Guides

• Sleep Training Methods: Ferber, No-Cry & More
• The Ferber Method: Step-by-Step Guide for Baby Sleep
• The Chair Method: A Gradual Approach to Sleep Training
• Sleep Training Age: When Can You Start?

Track your baby's sleep with Bebblo

Bebblo logs sleep, feeding, mood and notes with a single tap, keeping everything locally on your phone. During extinction training, seeing crying duration drop from 45 minutes on night one to 8 minutes on night five makes the hardest parenting week measurably worth it. Free, no mandatory account.

Track sleep training progress with Bebblo

The Weissbluth method moves fast — 1–3 nights — but tracking what happens on each of those nights gives you confidence that the method is working and helps you identify any patterns that need adjusting. Bebblo lets you log every sleep and wake with a single tap, giving you a clear night-by-night comparison showing bedtime settling time, night waking count, and total overnight sleep.

Start and stop the sleep timer, or enter the times manually. Bebblo uses those logged entries to estimate an age-based range for the next sleep. Since early bedtime is central to the Weissbluth approach, this loose estimate can support your timing alongside your baby's tired cues. Download free for iOS — no account required.

Comparing baby sleep apps? Read our full Bebblo vs Huckleberry breakdown.

Core tracking is free. No account required.