Why does everyone tell you to sleep train?
Is it because the evidence is overwhelming — so consistent and so clear that any reasonable person looking at the research would reach the same conclusion? Is it because paediatricians, midwives, and health visitors have independently evaluated the literature and determined that formal sleep training is the best approach for the majority of infants?
Or is something else going on?
The sleep training industry is not a conspiracy. There are sleep consultants who genuinely believe in their methods, and there are parents who have been genuinely helped. But it is a multi-billion-dollar sector with a financial model that depends on parents believing their baby’s sleep is a clinical problem requiring professional intervention. That financial model shapes what gets studied, what gets published, what gets recommended, and what you’ve been told.
Here’s what’s behind the certainty. For the full research picture, see our guide to cry-it-out. For permission to choose differently, read You Don’t Have to Sleep Train Your Baby. For the emotional side, see If Cry-It-Out Doesn’t Feel Right, Trust Your Instincts.
The scale of the industry
The global baby sleep industry encompasses books, apps, online courses, consulting packages, certification programmes for aspiring consultants, branded monitors, sleep suits, wake clocks, white noise machines, and subscription services. It is worth billions, growing annually, and it markets itself aggressively to a captive audience: parents in a state of chronic sleep deprivation who are desperate for a solution.
The financial model is straightforward. For it to work, parents need to believe three things: that their baby’s sleep is a problem; that the problem requires intervention; and that they are not capable of solving it themselves without professional guidance.
Night waking — which is developmentally normal for most babies in the first year, and for many babies well beyond it — gets reframed as a “sleep challenge.” Feeding to sleep — which is biologically designed, supported by the chronobiology of breast milk, and effective — gets reframed as a “negative sleep association.” A baby who needs a parent to settle between sleep cycles becomes a baby who “can’t self-soothe,” with the framing of deficiency rather than development.
What makes this particularly effective is the moment it arrives. The marketing enters the room at maximum vulnerability: the fourth trimester, the growth spurts, the periods of developmental change that disrupt sleep. A parent at three in the morning, running on fragments, is not in the best position to critically evaluate an industry’s evidence claims.
The commercial reach extends further than the obvious. Meta — the parent company of Facebook and Instagram — has subsidised sleep training programmes as a workplace productivity benefit for employees. (NPR Planet Money, 2024.) [5] The logic is transparent: a parent who is sleeping is a more productive worker. The interest in solving infant sleep is not only parental.
How certainty gets manufactured
The standard presentation of sleep training evidence goes roughly like this: “Research shows that sleep training is safe and effective.” This framing is not invented from nothing — there are studies that support parts of it. But it omits enough to be consistently misleading.
The evidence is presented as more settled than it is. The 2006 AASM review (Mindell et al.) — which found that extinction and graduated extinction showed effectiveness in 17/19 studies — is the most-cited source in pro-sleep-training literature [1]. It is real research. But it reviews studies that primarily measure parental reports of sleep problems. When objective measures are used — video monitoring, actigraphy — babies often continue waking at similar rates and simply stop signalling. A reduction in parent-reported wakings and a reduction in actual infant wakings are not the same thing [2].
Failure rates are omitted from the pitch. In real-world practice, 40–50% of parents cannot complete extinction approaches. In research settings, failure rates range from 16–50% [3]. If you tried sleep training and it didn’t work, you are in very substantial company — but you would not know that from the way the evidence is typically presented.
Long-term effects are inconclusive, not positive. “Sleep training has no long-term negative effects” is commonly presented as a benefit of the approach. What the evidence actually shows is no long-term difference — positive or negative — at five-year follow-up. The trained and untrained babies grew up with the same sleep quality, emotional development, and attachment outcomes. The intervention’s effects are time-limited [2].
Publication bias shapes what gets published. The overwhelming majority of published sleep training studies report positive outcomes. This is not because all sleep training research produces positive findings. It is because studies with positive outcomes are more likely to be written up, submitted, and accepted for publication. The studies that found no effect, or negative effects, are less visible. The published record is not a neutral representation of the evidence [2].
Normal behaviour is pathologised. The language of the sleep training world — “sleep debt,” “sleep crutch,” “negative sleep association,” “overtired spiral,” “resettling” — takes developmentally normal infant behaviour and reframes it as pathology requiring correction. Once night waking is a disorder, the solution is a product. Once feeding to sleep is a bad habit, the solution is a programme. The language is not neutral. It is doing commercial work.
The research bias
The imbalance runs deeper than publication bias. It shapes what questions get asked in the first place.
Extinction-based methods dominate the sleep training research literature not because they have been demonstrated to be superior to alternatives, but because they are easier to standardise and therefore easier to study. “Put the baby down and do not respond” is a protocol that can be applied consistently across a research population. “Respond to your baby in a gradually less intensive way over several weeks” is harder to operationalise, harder to control, and therefore underrepresented in the published record. The research base reflects research convenience — not clinical priority, and not necessarily what most families would choose if they had full information.
More revealingly: when sleep researchers identified parental distress as the main obstacle to extinction adherence, the research response was not to develop less distressing alternatives to extinction. It was to develop psychological techniques to help parents tolerate the distress that extinction was creating. Whittall et al. (2023) published work on “cognitive restructuring” — the mental reframing used to help parents override their instinct to respond to infant crying [4]. The goal of this research was to make extinction easier to complete by making parents more comfortable ignoring their baby. The alternative framing — that extinction might not be the right first-line approach — is less well-resourced.
60% of popular parenting books recommend CIO or controlled crying as a primary approach (Ramos and Youngclark, 2006) [6]. This figure predates the current online course, app, and social media ecosystem, which has significantly expanded both the reach and the rebranding.
Under six months, the evidence is particularly limited. Douglas and Hill (2013) found that early sleep interventions had minimal impact on outcomes and, in some cases, increased maternal anxiety and undermined breastfeeding [7]. The gap between “the evidence does not support this under six months” and “your health visitor has given this advice to a parent of a six-week-old” is large.
What this means for you
None of this means sleep training is wrong. It means the certainty with which it is sold is not warranted by the evidence base it claims.
Your scepticism — if you feel it — is rational. The gap between “this helps some families under some conditions” and “you need to do this for your baby’s development” is large, and it is filled with commercial interest, cultural assumption, and research that is narrower and more conditional than its popular presentation suggests.
If sleep training is right for your family — if you are comfortable with the method, your baby responds without excessive distress, and your household’s wellbeing improves — that is a valid choice backed by real evidence. Not everyone who sleeps trains is being manipulated. The approach genuinely helps many families.
But if it doesn’t feel right — if your instincts resist it, if you tried and couldn’t complete it, if the certainty around you doesn’t match your experience — you are not failing at a medical necessity. You are declining something that was never a medical necessity to begin with.
Your instinct to respond to your baby’s crying is a feature of the caregiving system. It is worth thinking carefully about who benefits from convincing you to override it.
The honest position
The sleep training industry sells certainty to exhausted people. Certainty that isn’t earned by the evidence is not clinical wisdom — it is marketing. And you deserve to know what the research actually says, including the failure rates, the objective measure gaps, the publication bias, and the absence of long-term benefit, alongside the genuine short-term benefits that exist.
Both things are true. The evidence is real and limited. The pressure is commercial and cultural. The choice is yours.
For the full research, see our guide to cry-it-out. For what to do if it’s not the right path for your family, read You Don’t Have to Sleep Train Your Baby.
References below.