Sleep

Is it OK to let baby cry it out?

S&M Sam & Mia · May 9, 2026

Short answer

After 4-6 months, yes — current evidence shows structured sleep training (including modified cry-it-out) does not harm attachment or cause long-term emotional damage, and it improves both baby and parent sleep. Before 4 months, no — the biology isn't ready and it doesn't work. The 'right' answer depends on baby age, parental tolerance, and whether both partners actually agree.

The current evidence (briefly)

The "cry-it-out" debate is one of the most-emotional topics in baby care, and most online discussion is heavier on opinion than data. The actual evidence base is fairly settled:

Gradisar et al. 2016 (the most-cited modern study) randomized 43 healthy infants aged 6-16 months to one of: graduated extinction (modified cry-it-out), bedtime fading, or a control group. At 12 months follow-up, children in the active groups had better sleep, no difference in attachment security or behavior outcomes, and parents had less depression and stress. Bilgin et al. 2020 replicated similar findings at 18-month follow-up.

What the evidence does not show: any short- or medium-term harm to attachment, cortisol regulation, or behavior. Older claims of "toxic stress" from sleep training have not been replicated in studies actually designed to test them.

This evidence applies to healthy babies aged 4-6 months and older, with structured methods (graduated extinction, Ferber, bedtime fading), with parents who are aligned on the approach. Outside those conditions, the evidence is mixed or absent.

What "cry it out" actually means

The term covers several different methods, and people often argue about "crying it out" while picturing different things.

Full extinction ("original Weissbluth"). Put baby down awake at bedtime. Don't return until morning. This is the strict version. Most modern sleep-training advocates do not recommend it; it's stressful for parents and the gentler methods work as well.

Graduated extinction ("Ferber"). Put baby down awake. If they cry, return at intervals (3 min → 5 min → 10 min → 10 min...) for brief comfort visits — pat, no picking up, leave again. The intervals lengthen each night. Most couples reach a settled pattern in 3-7 nights.

Chair method. Sit in a chair next to the crib until baby falls asleep. Each night, move the chair further from the crib. Takes 1-3 weeks. Lower-cry but slower.

Bedtime fading. Move bedtime later (toward natural sleep onset), then gradually shift earlier as baby learns to fall asleep without prolonged crying. Lowest-cry method; takes longest.

Pick-up-put-down. When baby cries, pick up briefly, then put back down once calm. Repeat. Effective for some babies; exhausting for parents; doesn't work for high-stim babies.

When people online say "don't do CIO," they usually mean full extinction. The graduated and chair methods are much less stressful and the evidence supports them more strongly.

When it works

Sleep training works reliably when:

1. Baby is at least 4-6 months adjusted age. Earlier than that, the biological prerequisites for sleep consolidation aren't in place. Trying to sleep-train an 8-week-old is like trying to potty-train a 6-month-old — the equipment isn't ready. 2. Baby is healthy. Not actively sick, not teething hard, not in a developmental leap. Pause for any of these. 3. Both partners are aligned. Partial buy-in does not work. If one partner is going in to comfort baby every 5 minutes while the other is following the protocol, the inconsistency makes the process longer and harder. 4. The bedtime routine is consistent. Bath → book → song → in crib at the same time every night. Sleep training works on a stable foundation; a chaotic bedtime is the actual problem you should fix first. 5. You're committed to a method, not switching after 2 nights. The first 2-3 nights are the hardest. Couples who switch methods at 48 hours never get the consolidation benefit.

When it doesn't work / shouldn't be tried

Under 4 months adjusted age — biology not ready
Babies with significant medical issues (severe reflux, food allergies causing pain, ongoing illness)
Babies with disrupted sleep environments (frequent moves, irregular schedules, traveling)
When parents are not aligned and it's becoming a fight
During major life disruptions (move, new sibling, parental illness, major schedule change)
If either parent is so emotionally distressed by hearing the baby cry that they cannot follow through. Forcing yourself through it produces worse outcomes than not doing it.

The couple part

This is where most sleep-training failures actually happen. The pattern: one partner reads a book or follows a protocol; the other partner instinctively responds to crying; the baby learns that crying produces the responder; the other partner gets frustrated; the marriage gets weird.

If you're considering sleep training, before you start:

1. Both read or watch the same source material. Pick one method, agree on the source, get on the same page about the rules. The most-recommended starting place: Solve Your Child's Sleep Problems (Ferber) or Precious Little Sleep (Alexis Dubief) — both are widely used and aligned with the modern evidence. 2. Decide who is on duty each night. One parent does the protocol that night — the other doesn't go in. Sharing duties across nights, not within a single night. 3. Set the abort criteria in advance. "If she cries for more than 30 minutes total in any one wake-up, we go in and the night is over." Knowing the criteria removes the in-the-moment fight about what to do. 4. Plan for setbacks. First 2 nights are the worst. Night 3 is usually a lot better. Night 7 is when things consolidate. If you don't see consolidation by night 10, the method probably isn't working — pause and reassess. 5. Don't try this during a hard week at work for either partner. The exhausted version of you cannot successfully sleep-train.

Alternative: the no-cry approach

The Pantley/no-cry methods are real and the evidence base for them is mixed. They work for some babies and don't for others. Generally:

They take 3-6 weeks instead of 1-2
They preserve breastfeeding-on-demand more easily
They don't work as well for babies with very strong feed-to-sleep associations
They require both partners to be patient on a longer timeline

If both partners deeply oppose any structured cry-period, the no-cry methods are the right choice. Just expect a longer arc to a consolidated night.

Bottom line

For most healthy babies 4-6 months and older, a structured sleep training method does not harm them and substantially improves family sleep. The decision of whether to do it is mostly about parental tolerance and alignment, not about whether it's "safe." The decision of which method to use is about how much crying you can tolerate weighted against how fast you need consolidation.

What matters more than the method choice: that both partners agree on it, do it consistently for at least 7-10 nights, and avoid switching mid-stream. Inconsistency is what creates the harder, longer process — not the method itself.