Gentle Sleep Training Methods Compared: What the Research Supports at Each Age

Every method has a name, a book and a group of parents who will tell you it saved their lives. The research is quieter. The most recent systematic review of the whole field found no settling method to be clearly superior to any other, and only a handful of approaches with enough trial evidence to call them empirically supported. Here is what each one is, what the evidence says, and who it fits, without the branding.

First, the frame

Three things are true of every approach below.

None of them should be used before six months, corrected for prematurity. Not because they have been shown to harm younger babies, but because they have never been tested on them and the sleep physiology is still being installed. Among parents who tried a checks-based method on their own, most started before six months and reported lower success.

None of them will make your baby sleep meaningfully more. Pooled across trials, behavioural approaches add about nine minutes of night sleep. What they reliably change is how long settling takes, how much of the night you are up attending, and how you feel. They also do not reduce how often your baby wakes; every baby wakes several times a night. They reduce how often your baby calls for you.

And the method a family can do consistently beats the method that is theoretically best. Ask yourself honestly how you feel about crying before you choose, because low tolerance for it predicts poor results with any waiting-based method, and that is fine. There are methods for you.

The foundations, at every age

Before any method, and instead of any method under six months: a bedtime routine and sensible timing.

The routine is the same two or three calm things, in the same order, twenty to thirty minutes, ending in the sleep space, every night. It is the single best-supported change in infant sleep, with a randomised trial of 405 families behind it and most of the benefit showing within three nights.

Timing means bright light by day and real darkness at night, and a bedtime that matches when your baby is actually sleepy. From about five months the morning wake time barely moves, so a bedtime that is too late subtracts sleep, and one that is too early produces resistance.

Your own protected block of sleep belongs here too. It is often the fastest thing a plan can change.

Bedtime fading

What it is. Temporarily set bedtime twenty to thirty minutes later than the time your baby actually falls asleep now, so they go down properly sleepy and are asleep within about fifteen minutes. Hold that for a few nights. Then bring bedtime earlier by ten to fifteen minutes every two or three nights, as long as settling stays quick. Keep the morning wake time fixed and do not let naps drift later to compensate.

What the evidence says. In a small randomised trial it produced settling gains comparable to graduated checks, with less crying and the largest improvement in mothers’ mood. The review classes it as promising but not yet well enough evidenced. No crying is built into it.

Who it fits. Almost anyone from six months, and toddlers especially, where bedtime resistance is very often a timing problem dressed up as defiance.

Fading the sleep-onset association

What it is. How a baby falls asleep at bedtime is one of the most reliable predictors of how much they signal overnight, so this approach moves, in small steps, from “feeds fully to sleep” to “feeds until drowsy, then transferred” to “into the crib drowsy, hand on chest” to “into the crib awake, parent nearby.” Each step is held for several nights before the next. It pairs with moving the feed earlier in the routine.

What the evidence says. The association itself is well documented; the reverse explanation (harder-to-settle babies get more help) is also live. As a method it is gentle, slow, and widely used rather than trial-proven on its own.

Who it fits. Any age as a direction of travel, and especially families who are happy feeding to sleep now but would like a way out later. Feeding to sleep at six to twenty-four weeks is the norm, not a problem to fix.

Gradual withdrawal, camping out, the chair method

What it is. You settle your baby in the crib and stay in the room, on a chair or a mattress beside it, offering a calm voice and brief touch as needed but not picking up to sleep. Every few nights the chair moves further from the crib, then to the door, then outside. Typically one to three weeks.

What the evidence says. Used in trials as the alternative to checks-based methods, and in one study babies who were highly distressed by separation did better with this than with being left. The review classes it as insufficiently evidenced on its own, but it is a mainstream, well-tolerated option.

Who it fits. Babies who are very distressed by separation, and parents who cannot leave the room and do not want to.

Graduated checks (controlled comforting)

What it is. After the routine, your baby goes down awake and you leave, returning at set, gradually lengthening intervals, for example two, four and six minutes on the first night and a little longer each night, to reassure briefly with a calm voice or a hand on the chest, without picking up. Applied at bedtime and at night wakings that are not feeds. Most families see clear change within three to seven nights.

What the evidence says. This is the most-studied class of method, with randomised trials, a five-year follow-up that found no lasting harm and no lasting benefit, and an attachment study at eighteen months that found no adverse effect. What has never been done is the stress-hormone study with a proper control group. So “harmful” is not supported by the research that exists, and “proven safe” overstates it.

Who it fits. Families whose tolerance for crying and whose values fit it. Some crying is part of this method; it is a legitimate choice for a family who chooses it, and not a requirement. Never under six months, never with a weight-gain concern, and paused for illness.

Scheduled awakenings

What it is. For a waking or a sleep terror that happens at a predictable time, you gently rouse the child fifteen to thirty minutes before it, for a week or so, then stop.

What the evidence says. Modest, and classed as insufficiently evidenced. Reasonable for stereotyped sleep terrors, where sleep deprivation is a trigger. Not for feeds, not for young infants.

What is appropriate at each age

• Under six months (corrected): foundations only. Routine, light and dark, feeding sorted, safe sleep every sleep, your own rest. Revisit at six months.
• Six to twelve months: every method above has its trial evidence from this age up. Match it to the baby: stay-in-the-room for the separation-distressed, fading for the low-crying-tolerance household, checks for the family that chooses them. Night feeds stay unless the pediatrician is happy with growth and you want to reduce them, and then gradually, never abruptly.
• Twelve to twenty-four months: the same menu, plus timing before technique. This age band is the centre of the behavioural treatment literature, and bedtime resistance is very often a too-early bedtime. Protect naps; fewer than one toddler in forty stops napping before two.

When to pause any method

Illness, a new red flag (snoring, breathing pauses, unusual movements, poor weight gain), a weight concern of any kind, travel, or your own exhaustion are all reasons to stop and go back to foundations. Plans are meant to be adjusted, and a bad night after a run of good ones is the expected shape of infant sleep, not a sign you undid something.

If you want the method chosen for your baby and written out night by night, that is what a Lamplight plan does, from your answers about your baby and about crying, for $79.