If you've tried sleep training with your autistic child and it didn't work — or made things worse — you are not alone. And you haven't failed.

Sleep training is one of the most commonly recommended approaches for children's sleep difficulties. It's backed by decades of research, endorsed by paediatricians, and fills entire sections of parenting bookshops. For many families with neurotypical children, it works well.

But for autistic children, sleep training often falls short. Sometimes it fails quietly — the child never quite settles the way the method promises. Sometimes it fails loudly — with escalating distress, increased night waking, or a child who is more dysregulated than before.

Understanding why sleep training doesn't work for autistic children isn't about criticising the approach. It's about recognising that it was designed for a different kind of nervous system.

Most sleep training approaches — whether gradual retreat, controlled comforting, or more structured methods — are built around the same core idea: that children have the neurological capacity to self-settle, and that the right conditions and consistency will allow that capacity to develop.

This idea is grounded in neurotypical sleep research. And for neurotypical children, it is often true.

The problem is that autistic children's nervous systems work differently. Applying a neurotypical framework to a neurodiverse nervous system is a bit like following directions designed for a different route. The destination might be the same. But the road doesn't match.

Research consistently shows that many autistic children produce melatonin later in the evening than neurotypical children. Melatonin is the hormone that signals to the brain that sleep is coming. When melatonin production is delayed, the biological readiness for sleep is delayed too.

This means that a bedtime set at 7:30pm may simply be too early for an autistic child's nervous system to be ready for sleep — regardless of how consistent the routine is. Sleep training at a time when the child isn't biologically ready for sleep will almost always produce resistance. Not because the method is being applied incorrectly, but because the body isn't ready.

Self-settling — the ability to move from wakefulness to sleep independently — requires the nervous system to be in a particular state. It needs to be calm, safe, and regulated.

Many autistic children arrive at bedtime carrying a significant regulatory load from the day. The sensory demands of school, the social effort of masking, the emotional labour of navigating a world that wasn't designed for their nervous system — all of this accumulates. By bedtime, many autistic children are running at a level of nervous system arousal that makes self-settling genuinely difficult, not because they won't, but because their body can't get there from here.

Sleep training approaches that ask a child to settle in this state are working against the nervous system rather than with it.

For autistic children with sensory processing differences, the bedroom environment can be full of sensory experiences that keep the nervous system alert. Sound, light, texture, temperature, smell — any of these can be processed more intensely than a neurotypical nervous system would process them.

Controlled comforting approaches, for example, ask a parent to leave the room and return at intervals. But if the child's distress is being driven by genuine sensory discomfort in the environment, leaving and returning at intervals doesn't address what's actually causing the difficulty. The sensory experience doesn't change because a parent comes back briefly. The nervous system stays activated.

Many autistic children need the regulatory support of a calm, present adult in order to bring their nervous system to a state that allows sleep. This is sometimes framed as a bad habit — the child has "learned" to need a parent present. But co-regulation is a neurological process, not a behavioural one. A dysregulated nervous system genuinely finds it easier to settle in the presence of a regulated other.

Sleep training approaches that systematically remove parental presence can feel destabilising to a nervous system that genuinely relies on that presence for regulation. This is why some autistic children become more distressed, not less, when sleep training is applied.

For many autistic children, transitions are a significant challenge. The transition to sleep — from wakefulness to unconsciousness, from the known world to the unknown territory of sleep — is one of the biggest transitions of the day. Anxiety around this transition is common and real. Sleep training approaches that don't address the transition anxiety are working around the problem rather than through it.

This isn't about dismissing sleep support as hopeless for autistic children. Sleep can improve, and it often does significantly with the right approach. The key is an approach that works with the autistic nervous system rather than asking it to behave like a neurotypical one.

Nervous system regulation across the whole day

Because the state the nervous system arrives in at bedtime is shaped by everything that has happened since waking, effective sleep support for autistic children usually begins well before bedtime. Building regulation support across the day — through sensory strategies, movement, predictable routines, and adequate decompression time after school — changes what the nervous system brings to the evening.

Sensory-informed environment design

Rather than assuming the bedroom is fine, an occupational therapist trained in sensory processing works through the sleep environment systematically. What does the child actually experience in that room? What sensory factors might be keeping the nervous system alert? Small, targeted changes here can make a meaningful difference.

Working with the child's natural rhythm

Rather than imposing a bedtime, working with the child's natural melatonin rhythm — even if that means a temporarily later bedtime — often produces faster progress than fighting the biology. Over time, the sleep window can be gradually adjusted if needed.

Proprioceptive input before bed

Heavy work activities in the wind-down period — activities that provide deep physical input to muscles and joints — have a calming, organising effect on the nervous system. Building these into the hour before bed can help lower arousal to a level where sleep becomes more accessible.

Graduated support rather than withdrawal

Rather than removing parental presence as a first step, an occupational therapist works to build the child's capacity to self-regulate gradually — while providing the co-regulatory support the nervous system genuinely needs in the meantime. This is a slower process than sleep training, but it tends to produce changes that hold.

If you've tried sleep training and it hasn't worked, or if you've decided it's not the right approach for your autistic child, neither of those things reflects badly on you or your child.

Sleep difficulty in autistic children is common, it has neurological roots, and it deserves support that's actually designed for how autistic nervous systems work.

That support exists. And it can make a real difference.

Jessinta Benton is a paediatric occupational therapist with a clinical focus on sleep for neurodiverse children. Dream Sleep OT is based in Maylands, Perth, and supports families across Australia via telehealth. NDIS self and plan-managed participants welcome. No referral needed.

Apply to work with Jessinta