Sleep training teaches your baby to fall asleep independently and stay asleep through the night, and most babies are ready to start at around 4 months old. There are several proven methods ranging from gradual to more direct, and the one you choose depends on how much crying you’re comfortable with and how quickly you want results. Here’s what actually works, how each method plays out night by night, and what to expect along the way.
Why 4 Months Is the Starting Line
Before about 3 months, babies don’t produce their own melatonin or regulate their own sleep cycles. Newborns sleep 16 to 17 hours a day in short 3- to 4-hour stretches scattered across day and night, and they genuinely need to eat during those wake-ups. They also haven’t developed the ability to calm themselves back to sleep without help.
Around 4 months, three things change at once. Your baby’s circadian rhythm kicks in, meaning their brain starts distinguishing day from night. Their sleep cycles begin to mature, with less active (REM) sleep and more deep, quiet sleep. And they become physically capable of self-soothing, which is the core skill sleep training builds. A general guideline from pediatric sources is that babies who weigh at least 14 pounds may no longer need overnight feedings, though your pediatrician can confirm this for your specific baby before you start.
Set Up the Room First
Sleep training works best when the environment supports it. Keep the nursery between 68°F and 70°F, with humidity between 30% and 50%. In summer, aim for the lower end of that temperature range; in winter, the upper end, since heating systems dry out the air. Use blackout curtains to block light, position the crib away from windows and heating or cooling vents, and keep the door slightly open for airflow. A fan pointed away from the baby can help with circulation and provides consistent white noise.
Every sleep training method starts with the same foundation: a predictable bedtime routine. Bath, feeding, book, song, whatever sequence you choose. The routine signals to your baby’s brain that sleep is coming. Then you place them in the crib drowsy but still awake. That last part is non-negotiable across all methods. If your baby falls asleep in your arms and wakes up alone in the crib, they haven’t learned to fall asleep independently, which is the entire point.
The Ferber Method (Graduated Check-Ins)
This is the most widely recognized approach. You put your baby down awake, leave the room, and wait a set number of minutes before going back in to briefly reassure them. You don’t pick them up. You might pat their chest, say a few soothing words, and leave again. Each time, the interval between check-ins gets longer. Each night, the starting interval gets longer too.
The schedule looks like this on the first three nights:
- Night 1: Wait 3 minutes before your first check-in, then 5 minutes, then 10 minutes between all remaining check-ins.
- Night 2: Start at 5 minutes, then 10, then 12 minutes for all subsequent check-ins.
- Night 3: Start at 10 minutes, then 12, then 15 minutes for all subsequent check-ins.
By night 7, you’re waiting 20 minutes before the first check-in and up to 30 minutes between later ones. These intervals come from Dr. Richard Ferber’s original protocol, though he notes in his book that parents can adjust them slightly. The key principle is consistent, progressive lengthening. Most families using this method see significant improvement within the first week, with many babies falling asleep independently by nights 3 through 5.
The hardest night is almost always the first or second. Crying often peaks early and then drops off sharply. If you find the intervals too aggressive, you can start with shorter waits (1 minute, 2 minutes, 3 minutes) and build from there. What matters is the pattern of increasing independence, not hitting exact numbers.
The Chair Method (Gradual Withdrawal)
If the Ferber method feels too intense, the chair method offers a slower path. After your bedtime routine, place your baby in the crib drowsy but awake, then sit in a chair right next to the crib. Stay there quietly until they fall asleep. Don’t pick them up, but your presence is the reassurance. If they wake and cry during the night, return to the chair and sit until they settle.
Every few nights, move the chair a little farther from the crib. First toward the middle of the room, then near the door, then just outside the door, then gone entirely. The idea is that your baby gradually adjusts to falling asleep with less and less of your presence, rather than all at once.
This method typically takes longer than Ferber, often two to three weeks rather than one. It also requires more patience in the moment, since you’re sitting quietly while your baby fusses, sometimes for 30 to 45 minutes in the early nights. Some parents find this harder than leaving the room because hearing the crying up close without responding feels unnatural. But for families who want to minimize crying or who have babies that escalate when left alone, it can be a better fit.
Full Extinction (Unmodified Crying)
This is the most direct method. You complete the bedtime routine, place your baby in the crib awake, say goodnight, and leave. You don’t return until morning (or until a scheduled feeding if your pediatrician says one is still needed). It sounds extreme, but it tends to produce the fastest results with the least total crying over the course of training, because check-ins can sometimes re-stimulate a baby who was close to settling.
A 2020 study tracking babies whose parents used this approach found no adverse effects on attachment or behavioral development at 18 months. Research on cortisol (the stress hormone) levels during sleep training has produced mixed results. One small study found elevated cortisol at the moment of separation, while others found no elevation when babies were allowed to settle on their own. The overall body of evidence suggests that sleep training, including full extinction, does not cause lasting emotional harm.
What Self-Soothing Actually Looks Like
The goal of every method is the same: teaching your baby to self-soothe. In practical terms, this means your baby can go from crying or fussing to quiet wakefulness and then to sleep without you intervening. It also means that when they naturally wake between sleep cycles during the night (which all humans do, including adults), they can put themselves back to sleep instead of crying for help.
During the first year, the percentage of time babies spend awake in their crib stays relatively constant, but the time parents spend getting them back to sleep drops dramatically between 1 and 3 months and then stays low. Sleep training accelerates this transition. You’re not teaching your baby something unnatural. You’re supporting a skill their brain is already developing.
How It Affects You, Not Just Your Baby
Sleep training is often framed as being about the baby, but the impact on parents is significant. In a large clinical trial, mothers who received sleep training guidance scored meaningfully better on depression screenings at both 10 and 12 months postpartum compared to mothers who didn’t. At 12 months, 25% of the sleep training group showed signs of clinical depression versus 28% in the control group, and their average depression scores were consistently lower across both time points. Their overall mental health scores were higher too.
This makes sense. Chronic sleep deprivation is one of the strongest predictors of postpartum depression and anxiety. When your baby sleeps through the night, you do too, and the downstream effects on mood, patience, and daily functioning are substantial.
Common Pitfalls That Derail Progress
Inconsistency is the biggest one. If you do Ferber for two nights, then bring your baby into bed on the third night because the crying is too hard, you’ve taught them that extended crying eventually works. Starting over after an inconsistent night is harder than the first night was, because your baby has learned that persistence pays off.
Timing matters too. Don’t start sleep training during a major disruption like travel, teething, illness, or a developmental leap. Pick a stretch of at least a week where your schedule is predictable and both caregivers are committed to the plan. If one parent handles bedtime differently than the other, the mixed signals slow everything down.
Watch for overtiredness. A baby who missed naps or stayed up too late gets a surge of cortisol that makes it harder to fall asleep, not easier. Counterintuitively, an earlier bedtime often produces less crying and faster sleep onset. If your baby is wired and frantic at bedtime, try moving it 30 minutes earlier.
Finally, expect regression. Around 8 to 10 months, separation anxiety peaks and many sleep-trained babies start protesting bedtime again. This is normal. Return to your chosen method for a few nights and the regression usually resolves quickly, because the underlying skill is already there.