Sleep Training Methods Compared, Honestly

You've read six articles and each one tells you to do something different. One promises results in three nights. Another says any crying will damage your baby forever. A friend swears by the chair. Your mother-in-law never heard of any of this and everyone in her generation turned out fine. Meanwhile the baby is awake again and it's 2:47 AM.

Here is the short version, ahead of the details: most established sleep training methods work if the family stays consistent for about two weeks, and the research does not show lasting harm from the extinction-based approaches when they are done after 6 months of age.¹ The best method is the one your household can actually run the same way for fourteen nights in a row. Pick for sustainability, not purity.

When is a baby old enough to sleep train?

Most pediatric sleep researchers set the floor at 4 to 6 months, with 6 months as the safer anchor for extinction-based methods.² Before 4 months, babies have not finished reorganizing their sleep architecture and genuinely need overnight feeds. If you are still in that window, your baby is not "failing" at sleep, they are developing on schedule. The 4-month sleep regression post covers what is normal at that age.

Three conditions worth meeting before you start:

• Baby is at least 4 to 6 months old and has been cleared by your pediatrician (no reflux flare, no ear infection, gaining weight on their curve).
• Bedtime and wake windows are already reasonable. Training a chronically overtired baby is training against their biology. See the bedtime guide and wake windows guide first.
• Both parents agree on the plan. The fastest way to stretch a 5-day method into 5 weeks is one parent giving in at 1 AM while the other is trying to hold the line.

Timing also matters within the week. Start on a night when you can tolerate a rough next day. Friday night is classic. Avoid starting during travel, a move, teething peaks, or the week a sibling starts a new school.

Graduated extinction (the Ferber method)

You put baby down drowsy but awake, leave the room, and check in at increasing intervals (typically 3, 5, 10 minutes on night one, longer on subsequent nights). Checks are brief, calm, and do not involve picking up.

• Age range: 6 months and up. Some consultants start at 4 to 5 months with shorter first intervals.
• Typical timeline: noticeable improvement by night 3, consolidated by nights 7 to 10.
• Evidence: Graduated extinction has the strongest evidence base of any behavioral sleep intervention. The 2006 AASM systematic review of 52 studies rated it as one of two approaches with robust empirical support, alongside full extinction.³ Gradisar 2016 found no differences in cortisol, attachment, or emotional/behavioral outcomes twelve months out compared to controls.¹
• Tradeoffs: There will be crying. Most families report it is hardest on nights 1 and 2, and the total crying time drops sharply by night 4. Parents who struggle to wait out the intervals often bail at the wrong moment and accidentally teach the baby that longer crying eventually summons them.
• Suits: Families who want a method with clear rules, a visible timeline, and strong evidence. Works well when both parents can agree on the check intervals and stick to them.

Full extinction (cry it out)

You put baby down awake and do not return until morning (except for safety checks or scheduled feeds if baby still needs them). No interaction, no check-ins.

• Age range: 6 months and up. Not recommended earlier.
• Typical timeline: often faster than graduated extinction. Many families see resolution within 3 to 5 nights.
• Evidence: Same AASM review flagged unmodified extinction as one of the two best-supported approaches.³ The five-year follow-up of the Hiscock trial (Price et al. 2012) found no differences in child emotional/behavioral health, sleep quality, stress, or parent-child relationship at age 6 between the intervention and control groups.⁴
• Tradeoffs: It is the hardest version emotionally. A single "just one check" resets the learning curve, which is why some families do better with graduated intervals they can actually hold to. Also requires a monitor and a baseline level of trust that the baby is safe.
• Suits: Families who have tried gentler methods without success, or who believe a shorter, more intense period is easier than a longer gentler one. Households where both parents are genuinely aligned.

The chair method

You sit in a chair next to the crib while baby falls asleep. Every 2 to 3 nights, you move the chair further from the crib, until eventually you're outside the room. No picking up, minimal talking.

• Age range: 6 months and up, sometimes earlier for families who want a very gentle approach.
• Typical timeline: 10 to 21 days. It is deliberately slower.
• Evidence: Less direct research on this specific method, though it falls under the broader "parental presence fading" family that the AASM review classifies under behavioral interventions with some support.³
• Tradeoffs: Slow. Baby often cries anyway because you are visibly present but not intervening, which some babies find more confusing than either picking up or leaving. Some parents find sitting silently in a dark room for 30 to 60 minutes nightly harder than brief check-ins.
• Suits: Parents who cannot tolerate leaving the room while baby cries, or babies with separation anxiety who settle faster knowing a parent is visible.

Pick up, put down

Baby cries, you pick them up until calm, then put them down awake. Repeat as many times as it takes. No rocking to sleep in your arms.

• Age range: 3 to 7 months is the sweet spot. After about 8 months, the pick-up itself becomes too stimulating and the method tends to stop working.
• Typical timeline: 1 to 3 weeks, with some long nights early on.
• Evidence: Limited direct trials. The Tracy Hogg approach that popularized it is not strongly evidence-based, though the underlying principle (separate falling asleep from parental contact gradually) aligns with what works in other methods.
• Tradeoffs: Exhausting for parents. Some babies get genuinely more upset from the repeated lift-and-place cycle and the method fails them. For others it is the only approach they can tolerate emotionally.
• Suits: Younger infants (4 to 6 months) where full extinction feels too much, and parents who want active involvement rather than waiting out intervals.

Bedtime fading

Instead of training the falling-asleep skill directly, you push bedtime later to a time baby is genuinely drowsy, let them fall asleep quickly with minimal crying, then gradually shift bedtime earlier by 15 minutes every few nights.

• Age range: 4 months and up. Also useful for toddlers with bedtime resistance.
• Typical timeline: 2 to 4 weeks to reach target bedtime.
• Evidence: Gradisar 2016 tested bedtime fading head-to-head against graduated extinction. Both worked, both showed no adverse stress or attachment effects, and fading had slightly less early-night crying.¹
• Tradeoffs: Slow. Requires tracking actual sleep-onset time carefully. Does not address middle-of-night wakings directly, so some families pair it with another approach for those.
• Suits: Families who cannot handle any protracted crying, babies with a history of bedtime battles, or households where the current bedtime is genuinely too early for the baby's sleep pressure.

No-cry and Pantley-style methods

The gentle umbrella: feed or rock to drowsy, put down, repeat gradually with increasingly hands-off support. Popularized by Elizabeth Pantley's "The No-Cry Sleep Solution." May include shortened nursing, a "pantley pull-off" to break the suck-to-sleep association, and incremental distance between parent and baby.

• Age range: Any age, including under 4 months where other methods are not appropriate.
• Typical timeline: measured in months, not weeks. Progress is often non-linear.
• Evidence: Minimal peer-reviewed data. Parent surveys suggest many families do see improvement but slowly, and it is hard to isolate the effect of the method from normal developmental maturation over the same window.
• Tradeoffs: Slow to the point that some babies outgrow the problem before the method resolves it. Not a failure mode if the family is content with gradual progress. Can become a trap if one parent is exhausted and waiting for faster results that are not coming.
• Suits: Parents who are philosophically opposed to any crying-it-out, or who are still bed-sharing and want to preserve that arrangement. Babies under 4 months where extinction-based methods are not yet appropriate.

Is sleep training safe? What the research actually says

The most cited safety concern is that crying causes a cortisol spike and that repeated spikes harm long-term development or attachment. The direct evidence does not support this at the ages these methods are used.

Gradisar et al. 2016 randomized 43 infants aged 6 to 16 months to graduated extinction, bedtime fading, or a control group. They measured morning and afternoon salivary cortisol during intervention, maternal mood and stress, infant emotional/behavioral problems at 12 months post-intervention, and parent-child attachment via the strange situation procedure. Both active methods improved sleep. Neither showed adverse cortisol, attachment, or emotional/behavioral effects compared to controls.¹

Price et al. 2012 followed 225 children from the original Hiscock behavioral sleep intervention trial to age 6. They assessed child mental health (child behavior checklist, parent-reported depression/anxiety), sleep, stress (salivary cortisol), and parent-child relationship. No differences between intervention and control groups on any measure.⁴ The Hiscock 2007 trial itself showed benefits for maternal mental health, with the intervention group reporting lower depression scores.⁵

The AASM 2006 systematic review of 52 studies concluded that behavioral interventions produce reliable, durable improvements in infant sleep, with 80% of treated children showing clinically significant gains maintained at 3 to 6 months.³

This research does not mean every method is right for every family. It means the choice between methods is about fit, not about which one is "safe."

How to pick a method that will actually work

Four questions to ask yourselves before night one:

1. How much crying can we tolerate on night 2? Not in the abstract, but specifically when the clock reads 11:47 PM and baby is 40 minutes in. If the answer is "almost none," do not pick full extinction.
2. Can we both run the same script? Write down the exact plan: put-down time, check intervals or fading schedule, who handles which wake. If either of you reads it and thinks "I'd probably just feed them," adjust the plan.
3. Is the current bedtime right? A wrong bedtime makes every method harder. Run through the bedtime guide for your baby's age first.
4. Are we ready for 14 nights? Most methods show most of their gains in the first week but need a second week to stabilize. Starting and stopping is worse than not starting.

When to pause or ask a pediatrician

Sleep training does not fix medical causes of poor sleep. Call your pediatrician before or during training if:

• Baby has a fever, persistent vomiting, or crying that sounds qualitatively different from protest.
• Weight gain has slowed or the pediatric curve is drifting.
• Sleep is worse after one week of consistent method, not better.
• There are signs of reflux, sleep apnea (pauses, gasping), or obvious discomfort.

Sleep training is a learning intervention. It only works if there is nothing physical standing in the way.

Frequently asked questions

Will sleep training hurt my baby's attachment to me?

The direct evidence does not show this. Gradisar 2016 measured attachment via the gold-standard strange situation procedure twelve months after sleep training and found no differences between graduated extinction, bedtime fading, and control groups.¹ Price 2012 followed children five years out with similar null findings.⁴

How long will my baby actually cry?

Highly variable, and mostly on nights 1 and 2. Most graduated extinction studies report total crying in the 20 to 45 minute range on night 1, dropping by roughly half each subsequent night. Full extinction often has a longer night 1 and shorter nights after. If crying is not trending down by night 4, something else is going on (wrong method for the baby, wrong wake windows, medical issue).

Can I sleep train while still nursing at night?

Yes. Most families keep one or two established night feeds during training and drop them separately later, or on a planned schedule. What training addresses is the non-feeding wakes, which at 6+ months are usually habit rather than hunger.

What if my baby starts standing in the crib and won't lie back down?

Common from about 8 months. Lay them down once, calmly, without conversation. Then follow your plan. Most babies stop pulling up for protest within 2 to 3 nights once they learn it doesn't change the response.