Drowsy But Awake: Why It Doesn't Work and What to Do Instead

You’ve been told this is the foundation of good sleep. Put your baby down drowsy but awake. Simple. Except your baby screams the moment their back touches the mattress. Or they fell deeply asleep at the breast and there is no drowsy window to find. Or you’ve tried it twenty times tonight and you’re both in tears.

Nothing. You are doing nothing wrong.

“Drowsy but awake” is the most widely repeated and least evidence-based piece of baby sleep advice in circulation. It is presented as the starting point — the entry-level move, so basic that any parent should be able to manage it. When it fails, the implication is that you’ve failed at the fundamentals.

You haven’t. The advice has.

This guide unpacks where the advice comes from, why it doesn’t work for most babies, what the research actually shows, and what to do instead. The short version: some babies are temperamentally suited to it and fall asleep this way easily. The majority aren’t. Neither group has anything to do with the quality of their parents’ technique.

For the brain science behind why babies can’t self-soothe on demand, see The Science of How Babies Fall Asleep. For the specific claims behind the advice, see The Myth of Drowsy But Awake. If you’ve tried and failed and need permission to stop, go straight to If ‘Drowsy But Awake’ Isn’t Working, You’re Not the Problem.

Where ‘drowsy but awake’ comes from

Drowsy-but-awake is a cornerstone concept of the sleep training industry. It is recommended by virtually every sleep consultant, parenting book, and paediatric sleep guide, and appears on NHS leaflets, in midwife discharge notes, and in the opening chapters of almost every baby sleep resource published in the last thirty years.

The underlying logic is intuitive: if a baby falls asleep in one place — your arms, the breast — and wakes in another — the cot — they will be confused and signal for help. If they fall asleep in the cot, they will be able to resettle there when they wake between sleep cycles. This is the sleep association framework: sleep onset conditions become learned associations, and babies who develop the “wrong” associations will be dependent on them indefinitely.

The couch-to-neighbour’s-house analogy is the vivid illustration most commonly used: imagine falling asleep on your sofa and waking up in a stranger’s house. You’d be confused. You’d call for help. Your baby, the logic goes, feels the same disorientation when they fall asleep in warm arms and wake in a cold cot.

The analogy is compelling. It is also invented. There is no research supporting the claim that babies experience spatial confusion of this kind, or that falling asleep in one location specifically causes distress upon waking in another. It is a plausible story that has been repeated so often it acquired the status of fact.

The framework originates from behaviourist sleep theory — the idea that sleep onset conditions are learned associations that can be modified through behavioural training. This approach largely ignores developmental readiness, treating the infant brain as a stimulus-response machine rather than a rapidly developing organ with built-in constraints. For more on how this framework shapes the wider “bad habits” conversation, see our guide to the bad habits narrative.

Why it doesn’t work for most babies

The first problem is the window itself. The advice asks parents to identify a precise neurological state — a specific point in the transition from wakefulness into sleep — and to act within it. Sleep professionals have called this the “unicorn sleep state” [4]: something discussed as if it is readily observable and reliably reproducible, but which many parents simply never manage to identify. The advice was built on the assumption that the window is universal. It isn’t.

For parents who do find a window, there is a second problem: placement. Most babies are being carried, rocked, or fed when drowsiness arrives. Transferring a drowsy baby to a flat, still surface requires a simultaneous change in temperature, movement, and touch. For many babies, this is enough to trigger the neurological alarm that snaps them back to full wakefulness.

This alarm system is not a behaviour problem. Young babies are wired to detect separation from their caregiver. When placed on a flat, motionless, solo surface after being held in warm arms, their nervous system registers a threat. Cortisol rises. Crying follows. This is adaptive, not defiant — a survival mechanism that kept infants close to their carers for most of human history.

Self-regulation isn’t a trainable skill

The deeper issue is developmental. Drowsy-but-awake rests on the assumption that babies can learn to fall asleep independently — that if you consistently put them down in the right state, the capacity will eventually establish itself. But falling asleep independently requires self-regulation: the ability to manage physiological arousal, reduce cortisol, and transition through sleep stages without external support.

Self-regulation is a brain maturation process, not a behaviour. It requires three systems to be functional and integrated: the prefrontal cortex (which governs planning and inhibition), the limbic system (which processes emotion), and the autonomic nervous system (which manages the stress response). In infants, all three are immature. The neural pathways required for genuine self-regulation may not be fully developed until well into childhood — and some research suggests key elements continue developing into puberty [4].

Expecting a four-month-old to self-soothe is not a reasonable behavioural goal. It is, as the developmental neuroscience shows, a request for a skill the brain does not yet possess.

Temperament matters

Some babies do fall asleep easily when placed down drowsy. Their parents often credit the technique. What they may be observing is temperament: an innately lower sensitivity to transitions, a lower arousal threshold, a constitutional ease with environmental change. This is not caused by technique. It is something those babies had from birth.

For temperamentally sensitive babies — those with higher arousal responses, stronger startle reflexes, or a greater need for proximity — no amount of technique refinement makes drowsy-but-awake smoother. The advice treats all babies as identical. They aren’t.

Breastfeeding biology works against it

For parents who are breastfeeding, the biology creates an additional obstacle. Evening breast milk contains cholecystokinin (CCK) — released in the baby’s gut during feeding, producing simultaneous satiety and sedation — as well as tryptophan (a melatonin precursor) and sleep-promoting nucleotides that are higher in evening milk than daytime milk. By the time an evening feed ends, the baby is often either fully asleep or snapping back to alertness as the breast is removed. The “drowsy but awake” window may not exist at all. For more, see our complete guide to feeding to sleep.

What the research actually shows

There is some research relevant to drowsy-but-awake. It is more limited, and more complicated, than the universal advice would suggest.

Goodlin-Jones et al. (2001) conducted a videosomnography study of 80 infants across four age groups. The researchers distinguished “self-soothers” — infants who returned to sleep independently after waking — from “signalers” — those who called for parental help. Two findings are consistently under-cited in the popular conversation: at 12 months, 50% of infants still typically needed parental help to return to sleep after waking. And self-soothers woke just as often as signalers — the difference was whether they vocalised, not whether they slept better [1].

Silence is not the same as sleep quality.

Burnham et al. (2002) followed 80 infants from birth to 12 months and found that the strongest predictor of self-soothing at 12 months was high levels of quiet sleep at birth — a constitutional characteristic, not a trained one [2]. Some infants simply don’t follow the “standard pathway” to self-soothing, regardless of parenting approach.

St James-Roberts et al. (2022) found associations between putting babies to bed awake and later self-soothing [3]. The sample was 20 families. The design was observational — correlation, not causation. And the study does not specifically validate the “drowsy” component; it looks at awake versus asleep at placement as a broader category.

None of this means drowsy-but-awake is entirely baseless. It means the evidence is too weak and too conditional to support the confident, universal way it is typically presented. If it works for your baby, that tells you something about your baby’s temperament. It does not validate the advice as a universal principle.

What to do instead

If drowsy-but-awake is working for you

Keep doing it. Some babies are genuinely amenable to it — they transfer easily, settle quickly, and thrive on the approach. If you’re in that group, none of this article concerns your parenting.

If it isn’t working: stop trying

The repeated failure-and-crying cycle — ten attempts a night, escalating distress, an increasingly anxious parent and an increasingly dysregulated baby — is measurably worse for everyone than simply helping your baby to sleep in a way that works.

You are not creating a problem. You are meeting a developmental need.

Feed your baby to sleep. Rock them. Hold them. Walk them in a carrier. Whatever works. None of these approaches creates a dependency that cannot be outgrown. All children eventually develop the capacity to fall asleep independently — it happens on a neurological timeline, not a technique schedule. See our guides to contact sleeping and feeding to sleep for the evidence behind these approaches.

Diversify sleep cues over time

Rather than obsessing over the specific state your baby is in at placement, focus on building a consistent sleep environment: a dark room, white noise, a reliable bedtime routine, appropriate wake windows for their age. Over time, introducing multiple sleep cues — a particular song, a specific phrase, a comforter once age-appropriate — means no single one is essential. This is a lower-stakes, more realistic goal than achieving the perfect drowsy window.

Trust the developmental timeline

Babies who need significant help to sleep at four months will, in most cases, need considerably less help at eight months, and less again at 12 months. This happens naturally, without training, as the brain matures and the capacity for self-regulation develops. It happens on a different schedule for every baby. Temperament and development are more powerful predictors of eventual independent sleep than any technique used in infancy.

If you want to actively encourage independent sleep

If, at six months or beyond, you want to move towards your baby needing less parental support at sleep onset, gentler approaches exist. Stay-with methods, gradual retreat, and responsive settling all work within the baby’s developmental capacity rather than against it. They do not require a specific drowsiness level at placement and allow both the baby and the parent to adjust at a pace that works.

Whatever path you choose, there is no single correct way for a baby to fall asleep. There is only what works for this baby, at this developmental stage, in this family.

A note on where this leaves you

Drowsy-but-awake has been treated as the absolute starting point of all baby sleep advice — the baseline that everything builds on. For many families, it is the starting point of sleep anxiety, self-doubt, and unnecessary guilt.

If it works for you, use it. If it doesn’t, put it down. Your baby will develop the capacity to fall asleep independently — when their brain is ready, in their own time, in whatever way suits their temperament.

For the full science, read The Science of How Babies Fall Asleep. For the claims behind the advice, see The Myth of Drowsy But Awake. For permission to stop, read If ‘Drowsy But Awake’ Isn’t Working, You’re Not the Problem.

---

References below.