Cry-It-Out: The Evidence vs the Internet

It's 3 a.m. and you've Googled "cry-it-out damage" with one hand while rocking a nine-month-old with the other. The results are terrifying: attachment disorders, cortisol spikes, babies who "learn that no one comes." Then you find the opposite — forums full of parents saying it was the best decision they ever made. Both camps are utterly certain. Here's what the actual research says, without the ideology.

What the Studies Actually Found

Two well-designed studies are worth knowing by name. Gradisar et al. (2016), published in Pediatrics, randomly assigned infants to extinction sleep training (what most people call cry-it-out), graduated extinction (Ferber-style), or a control group. At follow-up — measuring cortisol levels, infant-parent attachment security, and child emotional and behavioral development — there were no measurable differences between the groups. None. The babies who cried it out did not show elevated stress hormones the next morning, and their attachment to their parents looked identical to babies in the control group.

Price et al. (2012) followed children from infancy to age six and similarly found no evidence of lasting harm to mental health, behavior, or parent-child relationships in families who used behavioral sleep interventions. The authors were careful to note the limits of their data, but the finding held: crying during sleep training did not produce detectable damage at follow-up.

Neither study promises a perfect night's sleep or suits every family. What they do say, clearly, is that the harm narrative circulating online is not supported by the evidence we currently have.

Where the Internet Goes Wrong

The loudest voices online tend to come from two places: parents who had a genuinely hard experience and parents who had a genuinely easy one. Both are real. Neither is a controlled study.

A few specific claims worth fact-checking:

• "Babies learn their needs won't be met." Research on attachment (Bowlby, Ainsworth and decades of follow-up work) consistently shows that attachment security is built across thousands of daily interactions — feeding, comfort, play, eye contact — not determined by a few nights of crying at bedtime.
• "Cortisol stays elevated for days." Gradisar's data showed cortisol normalizing, not spiking, over the intervention period. The cortisol studies cited to scare parents typically involve prolonged, unresponsive neglect — a very different situation.
• "It only works temporarily." Some children do have sleep regressions after training, for developmental reasons unrelated to the method. This is true of every sleep approach.

None of this means extinction is painless for parents. Listening to your baby cry is genuinely hard, and that emotional cost is real and valid regardless of what the data show.

Why It Works for Some Families and Not Others

This is the part the internet tends to skip. Temperament matters — a lot. Research on infant temperament suggests that babies who score higher on intensity and lower on adaptability tend to cry longer and harder during extinction, which makes the process more distressing for everyone and increases the chance parents abandon it partway through (which can prolong the difficulty without achieving the goal).

Consistency also matters. Extinction works on the principle that a behavior that is never reinforced eventually extinguishes. Intermittent reinforcement — going in on night four because you simply can't take it — can actually make crying more persistent, not less. This isn't a moral failing; it's just how behavioral learning works, and it's useful to know going in.

Other factors that shape outcomes:

• Whether the baby has an underlying reason for waking (hunger, illness, ear infections)
• Whether both caregivers are on the same page
• Whether the family has enough support to get through the first few nights

Extinction is a parental choice, not an obligation — and it isn't appropriate for babies under roughly four to six months, whose neurological development and feeding needs make behavioral sleep training unsuitable. The AAP recommends waiting until at least this age before attempting any sleep training method.

If you're considering extinction, the single most practical next step is to spend one week logging your baby's natural sleep patterns — when they fall asleep, when they wake, how long resettling takes. That data helps you spot a workable bedtime window (most infants sleep best between 7 and 8 p.m.) and gives you a realistic baseline so you can actually tell whether things are improving. Whatever you decide, you're not choosing between loving your child and sleeping. You're choosing between methods, and you know your child's temperament better than any study does.