The Ferber method is a sleep training technique where you put your baby to bed awake and then check on them at gradually increasing intervals if they cry, rather than picking them up or staying in the room until they fall asleep. Developed by pediatric sleep researcher Dr. Richard Ferber, the approach teaches babies to fall asleep independently by giving them brief reassurance at timed intervals while progressively stretching the time between visits. Most families see significant improvement within three to seven nights.
How the Method Works
The core idea is simple: after a calming bedtime routine, you place your baby in the crib drowsy but still awake, then leave the room. If your baby cries, you wait a set number of minutes before going back in to briefly reassure them with your voice or a gentle pat. You don’t pick them up, turn on lights, or feed them. Each check-in lasts only a minute or two, just enough to let your baby know you’re still there.
The key feature is the graduated waiting. On the first night, you might wait 3 minutes before your first check, then 5 minutes, then 10 minutes, and continue at 10-minute intervals after that. Each subsequent night, you start with longer waits and stretch them further. By night three or four, initial waits might be 10 or 12 minutes, with maximums around 20 minutes. The intervals increase across the first week until most babies begin falling asleep before the first check-in is even needed.
When you do check in, the goal is minimal stimulation. Keep the lights off, speak softly, and avoid picking your baby up. You’re signaling that you’re nearby and that everything is safe, not restarting the process of getting them to sleep. If your baby wakes during the night, you follow the same interval schedule from where you left off.
When to Start
Babies don’t begin producing their own melatonin or developing regulated sleep cycles until around 3 months of age. Before that point, their ability to distinguish between day and night and connect sleep cycles is limited, so sleep training of any kind isn’t appropriate.
Most pediatric sleep experts recommend waiting until at least 4 months of age and 14 pounds of body weight. The weight threshold matters because babies at that size can typically go through the night without needing a feeding. By around 6 months, many babies are developmentally ready to self-soothe and fall back asleep on their own, making this a common starting point for the Ferber method specifically. If your baby was born premature, their adjusted age (not calendar age) is the relevant number.
Setting Up for Success
The method works best when the groundwork is already in place before you start the timed intervals. A few things matter more than most parents realize.
A predictable bedtime routine. A consistent sequence of steps every night, like a bath, a book, and a song, helps your baby’s brain recognize that sleep is coming. This priming effect is real: routine signals shift a baby’s nervous system toward winding down before you ever place them in the crib.
The right sleep environment. The room should be dark, quiet, and free from distractions. When you check in during intervals, keep lights off and avoid anything stimulating.
Daytime activity. Babies who get plenty of physical activity and stimulation during the day tend to sleep better at night. On the flip side, avoid screens or highly stimulating play in the hour before bed.
Drowsy but awake. This is the most important single instruction. If your baby falls asleep while being rocked, fed, or held, they haven’t practiced the skill the method is designed to teach. The entire point is for your baby to experience the transition from awake to asleep while lying in the crib.
What the First Few Nights Look Like
The first night is typically the hardest. Many babies cry for 30 to 60 minutes before falling asleep, and parents often find the waiting more difficult than their baby does. The second night can sometimes be worse, a phenomenon sleep consultants call the “extinction burst,” where crying temporarily intensifies before it drops off. This is normal and actually signals that the process is working.
By night three or four, most babies cry significantly less. By the end of the first week, many fall asleep within minutes of being placed in the crib, sometimes with no crying at all. The method tends to work fastest when parents are consistent. Giving in on night two and rocking your baby to sleep resets the learning process and can make the next attempt harder, because your baby learns that extended crying eventually gets the result they want.
Night wakings follow the same pattern. You’ll likely still have middle-of-the-night crying for the first few nights, but it drops off on roughly the same timeline as bedtime crying. If your baby is still at an age or weight where a nighttime feeding is appropriate, you can preserve one scheduled feed without undermining the method.
The Debate Over Stress and Attachment
The Ferber method is one of the most debated topics in parenting, and the concern centers on one question: does letting a baby cry cause lasting harm?
Critics point to the biology of stress. When a baby cries and doesn’t receive an immediate response, their body activates the fight-or-flight system, releasing cortisol. Sustained or frequent spikes in cortisol outside of normal patterns can, in theory, affect brain development over time. Research on parent-infant attachment has also shown that mothers who consistently and promptly respond to nighttime crying tend to have more securely attached infants. From this perspective, any method that delays response to crying carries a risk.
Supporters, including many pediatric sleep researchers, counter that graduated extinction is fundamentally different from ignoring a baby. The check-ins provide reassurance, the intervals are structured and temporary, and the process typically resolves within a week. Several studies on behavioral sleep training methods, including graduated extinction, have found no measurable differences in cortisol levels, emotional development, or parent-child attachment at follow-up assessments one, two, and even five years later. The American Academy of Sleep Medicine considers behavioral sleep interventions, including graduated extinction, to be effective and evidence-based.
The honest summary: short-term stress during sleep training is real and observable. Long-term harm from graduated extinction, when used at the appropriate age and with proper check-ins, has not been demonstrated in controlled studies. But the research isn’t infinite, and every baby is different. Some babies respond quickly and cry very little. Others are more persistent, and parents may find the process doesn’t feel right for their family.
When the Method Isn’t Working
If crying isn’t decreasing by night four or five, something else may be going on. The most common culprits are inconsistency (sometimes picking the baby up, sometimes not), an overtired or undertired baby whose schedule needs adjusting, or an underlying issue like teething, illness, or an ear infection that makes it painful to lie down.
Some babies also have temperaments that don’t respond well to graduated extinction. Highly persistent babies may escalate rather than wind down, and for these children, gentler approaches with more parental presence in the room may be a better fit. The Ferber method is one of several validated sleep training techniques. It’s the best known, but it isn’t the only option.
How It Differs From “Cry It Out”
Parents often use “Ferber method” and “cry it out” interchangeably, but they’re different approaches. Full extinction, or “cry it out,” means placing the baby in the crib and not returning until morning, no matter how long they cry. The Ferber method’s graduated check-ins are specifically designed to avoid that. You are present, repeatedly, throughout the process. The intervals simply get longer over time so your baby has the space to practice falling asleep without your direct involvement. For many parents, this distinction is what makes the Ferber method feel manageable when full extinction feels too extreme.