Few topics in parenting generate more heat, more guilt, and more contradictory advice than sleep training. You will find passionate advocates on both sides, studies cited to support completely opposing conclusions, and more than a few people who will tell you - with complete confidence - that you are doing permanent damage to your child either by doing it or by not doing it.
Most of them are wrong. Or at least, significantly overstating what the evidence actually shows. Here's what the research says, as clearly as possible, without an agenda.
What sleep training actually is
Sleep training is a broad term that covers a wide range of approaches - from extinction methods (often called cry it out, where you put the baby down and don't return until morning) to graduated approaches (returning at increasing intervals) to gentler methods (gradual withdrawal, pick-up-put-down, camping out). These are not the same thing, and the research doesn't treat them as equivalent - though popular discussion often does.
The thing they have in common: they all involve some degree of teaching a baby to fall asleep independently, rather than relying on a feed, a hold, or another external prop to get there.
What the evidence shows
The short version: sleep training works, in the sense that it reduces night wakings and improves sleep consolidation. Multiple randomised controlled trials show this. It also does not, based on current evidence, cause lasting psychological harm.
The most cited study - a 2012 Australian randomised controlled trial published in Pediatrics - followed children who had undergone graduated extinction and camping-out sleep training at five years old. It found no differences in emotional and behavioural outcomes, attachment security, or cortisol levels compared to control children. A 2016 follow-up found the same.
A 2023 review of 13 studies concluded that behavioural sleep interventions improve infant and maternal sleep without evidence of harm to infant emotional development, attachment, or stress responses.
What the evidence doesn't show
It doesn't show that sleep training is risk-free - the absence of evidence of harm is not the same as evidence of no harm, particularly for very young babies. Most studies are conducted on babies older than six months. The research on infants under six months is thinner.
It also doesn't show that sleep training is necessary, or that babies who aren't sleep trained are worse off. Children who are not sleep trained also grow into securely attached, well-functioning people. Sleep training is a tool - not a requirement.
The cortisol argument
You may have encountered the claim that sleep training causes cortisol spikes that damage the developing brain. This comes primarily from a 2012 study by Wendy Middlemiss, which found elevated cortisol in babies even after they stopped crying during sleep training. It was a small study, it has been criticised methodologically, and its findings have not been consistently replicated.
This does not mean infant stress responses are irrelevant. It means the science is more complicated and more contested than either side tends to acknowledge.
What this means practically
If your baby is over six months, you are exhausted to a degree that is affecting your mental and physical health, and you want to try sleep training - the evidence suggests it is likely to help and unlikely to cause lasting harm. You are not damaging your child.
If you don't want to sleep train - because it doesn't feel right for you, because your baby isn't ready, because you're not ready - the evidence does not suggest your child will suffer for it. You are not creating a rod for your own back.
The bottom line
The research supports parental autonomy here. Do what is sustainable for your family, informed by evidence rather than guilt.
"Sleep training works. It also does not, based on current evidence, cause lasting psychological harm."
"The absence of evidence of harm is not the same as evidence of no harm."
"Do what is sustainable for your family, informed by evidence rather than guilt."