Infant Sleep Training: What Actually Works at 4 to 24 Months

You are here because something that used to work stopped working. Maybe your baby went down at 7 and stayed down, and now you are doing three trips down the hall a night. Maybe naps used to be an hour and now they are 35 minutes on the dot. Maybe the whole thing is fine except your baby thinks the day starts at 5 in the morning and you have run out of ways to disagree.

Most infant sleep training advice hands you a method on the first page. I am not going to do that, because the method is the second decision. The first one is figuring out which problem you have. A baby who cannot fall asleep without you, a baby who is genuinely hungry at 2 in the morning, a baby who is short on total sleep, and a baby whose bedtime is in the wrong place all look identical at 3 a.m. and need four different responses. Pick the wrong one and you will do everything right for a week and get nothing.

So we are going to sort your nights first, then pick one thing, then hold it long enough to know whether it worked.

What infant sleep training actually means, and what it does not

Infant sleep training is a set of behavioral approaches for helping a baby fall asleep at the start of the night, and fall back asleep in the middle of it, without the specific thing you have been doing to get them there. That is the whole scope. It is not a treatment for reflux, teething, ear infections, or a growth spurt, and it will not outrun a genuine medical issue.

That matters for two reasons. First, if your baby's sleep changed suddenly and nothing about your routine changed, that is a pediatrician conversation before it is a method conversation. I cannot tell you what is going on with your baby and this guide is not medical advice for your situation. Second, sleep training only addresses the falling-asleep part, which means if your baby is short 2 hours of total sleep a day, teaching them to self-settle will not fill the gap. You have to fix the amount separately.

On the research side, the most commonly cited trial here is Gradisar and colleagues in Pediatrics in 2016, which randomized 225 infants to graduated extinction, bedtime fading, or a control group. Both active approaches got babies to sleep faster than the control condition, and at 12-month follow-up the researchers found no differences between groups in infant cortisol, child emotional or behavioral problems, or parent-child attachment. That is one trial, not a settled question, and it does not tell you which approach fits your family. It does tell you that the choice between the common methods is more about what you can actually execute than about which one is safe.

Before you change anything: the safe sleep floor

Whatever you decide about method, the sleep environment is not one of the variables. The American Academy of Pediatrics' 2022 recommendations in Pediatrics are explicit: infants sleep on their back, on a firm and flat surface, in their own crib, bassinet, or play yard, in the parents' room, with no soft bedding, pillows, blankets, or bumper pads. The AAP recommends room sharing for at least the first 6 months.

If you have been solving night wakings by bringing your baby into your bed, that is exactly where the safe sleep guidance and the desperation collide, and it is worth reading the AAP recommendations directly rather than taking my summary of them. Sort the sleep surface out before you start any plan, because otherwise you are building a routine on top of something you will have to change anyway.

Which wake-up are you actually getting?

Spend 3 nights writing things down before you change anything. Not an app, a notepad next to the crib. You want: time down, time asleep, every waking, what you did, how long it took, morning wake time, and every nap with its length. Three nights, because one night is noise and a week is too long to wait when you are this tired.

Then sort what you wrote into one of these.

The handoff wake-up. Your baby falls asleep on you, in the stroller, or on the bottle, and wakes 40 minutes later or at every sleep cycle boundary wanting the same conditions back. The tell is that the first stretch of the night looks fine and the back half falls apart, and that whatever puts them down works every single time it is applied. This is the one the classic methods are built for.

The hunger wake-up. One reliable waking at roughly the same time, a real feed rather than a few sucks, and then a clean return to sleep. Under 6 months especially, this is often just a feed your baby still needs. Night weaning is a conversation with your pediatrician, not a decision you make off a blog post, because it depends on weight gain and daytime intake.

The sleep-debt wake-up. Early mornings, 30-minute naps, and a baby who is hard to put down despite being clearly exhausted. Counterintuitively, a baby short on sleep sleeps worse, not more. Check your totals against the American Academy of Sleep Medicine consensus statement in the Journal of Clinical Sleep Medicine, which recommends 12 to 16 hours per 24 hours including naps for infants 4 to 12 months, and 11 to 14 hours for children 1 to 2 years. If you are meaningfully under that band across 24 hours, the amount is your problem, not the settling.

The schedule wake-up. Bedtime is not where your baby's body thinks it is. Too early and you get a long protest at bedtime plus a 5 a.m. start. Too late and you get an overtired, hard-to-settle baby who also wakes early. This is what the bedtime fading arm of the Gradisar trial was addressing: you move bedtime to when your baby is actually falling asleep, then walk it earlier in small steps once it consolidates.

One more note for the two-caregiver households, which is most of you reading this. Write down who did each wake-up. If one of you is a 4-minute settle and the other is a 40-minute settle, you do not have a baby problem, you have a consistency problem, and no method survives that.

Pick one method and stop reading about the others

Once you know which wake-up you have, the choice narrows fast.

If it is a handoff wake-up, you are choosing between graduated extinction, where you put your baby down awake and return at set intervals to check in briefly without recreating the sleep association, and a slower in-room approach where you stay present and reduce your involvement in steps across a week or more. Graduated extinction generally produces change faster. The in-room version asks more nights of you but is often the one that couples can actually agree on. Both were represented in the research above; neither is the correct answer for everyone.

If it is a schedule wake-up, you are doing bedtime fading, and you are not doing anything else at the same time. Move bedtime to the time your baby is actually falling asleep, hold it 3 nights, then shift 15 minutes earlier at a time once sleep onset is happening within 15 minutes of lights out.

If it is a sleep-debt wake-up, your work is on the daytime side: protect the naps you have, cap the late-afternoon one if it is eating into bedtime, and get bedtime early enough that you are inside the AASM band across 24 hours. Then reassess whether there is still a settling problem underneath.

If it is a hunger wake-up, you talk to your pediatrician first.

The rule underneath all of this is that you change one variable at a time. If you move bedtime, drop a feed, and start interval checks on the same night, you will not know what worked and you will not know what to keep.

How long before you decide it failed?

Give it 4 nights before you judge anything, and expect night 3 to be worse than night 2. That pattern is the single most common point where families quit, one night before the thing starts working.

What you are watching for across those 4 nights is direction, not perfection. Is time-to-sleep at bedtime trending down? Are the wake-ups fewer, or shorter, or both? A baby who went from 3 wake-ups to 2 and from 40 minutes of settling to 15 is responding, even though you still got up.

If 5 or 6 nights of genuinely consistent response produce no movement in any of those numbers, you probably misdiagnosed the wake-up. Go back to your notes. The usual miss is a schedule problem being treated as a settling problem, or a sleep-debt problem hiding underneath both.

And plan for the fact that this will come apart again. Teeth, a cold, travel, a nap transition, the week your baby learns to pull to stand. Something disrupts it roughly every few weeks, and every time it does you will run the same loop: 3 nights of notes, sort the wake-up, change one thing, hold 4 nights. The loop does not get harder. You just get faster at it, because you already know what your baby's four patterns look like on paper.

What to have written down before night 1

By the time you turn the light off tonight, you want five things decided and, ideally, on paper where the other adult in the house can see them.

The wake-up pattern you have identified from your 3 nights of notes. The one method you picked, named, with the exact interval or step schedule. Bedtime, with a separate earlier bedtime for daycare days. Who is on which wake-up, so nobody improvises at 2 a.m. And the specific numbers you will look at on the morning of night 5 to decide whether this is working.

That last one is the part people skip, and it is why so many families cannot tell you whether the last attempt worked. Write the numbers down before you start. Tired you, 4 nights from now, is not a reliable narrator of how the week went.

You do not need to get this right the first time. You need a plan specific enough to be wrong in a way you can see, so that when you adjust it you are adjusting something real.