Translated from the Portuguese original. Read the original (Português)
The Early Start Denver Model (ESDM) is a program for very young autistic children — roughly 12 to 48 months old. Its hallmark is teaching within play: instead of sitting the child at a table to drill isolated skills, the adult joins the activity the child has already chosen and embeds the learning objectives there. It combines principles of behavior analysis with what is known about child development and relationships. On results, honesty matters: there is evidence of gains in language and cognition, of moderate magnitude, but the broad gains from the original study were not fully replicated — and the model does not "take anyone off the spectrum".
The ESDM is a comprehensive early intervention model for young children with autism, developed by Sally Rogers and Geraldine Dawson and manualized in 2010. It belongs to the family of naturalistic developmental behavioral interventions (NDBI): it applies the principles of behavior analysis — antecedent, behavior, consequence, reinforcement, trial-based teaching — within natural play and caregiving routines, guided by the child's interest, with objectives set by its own developmental curriculum and progress tracked through data collection and implementation fidelity checks.
What makes Denver different
The question the model answers is not "what to teach," but "where to teach." Its bet is that a two-year-old learns more in play they enjoy than in a sequence of tasks chosen by an adult — and that social interaction itself, when it's fun, becomes the reinforcer. Tap the tabs:
Teaching fits inside play
Objectives aren't drilled separately: they're embedded in activities that already make sense to the child — stacking blocks, blowing soap bubbles, bath time, eating. Within a single play routine, joint attention, imitation, requesting, vocabulary, and turn-taking can all be targeted at once, without any of it feeling like "class" to the child.
The child initiates, the adult joins in
The starting point is the child's initiative, not the adult's agenda. The therapist observes what the child sought out, joins it as a play partner, and creates learning opportunities from inside the activity. This changes the tone of the session: the model gives explicit weight to positive affect and reciprocity — smiling, taking turns, having fun together are part of the procedure, not decoration.
No improvising
Looking like loose play doesn't mean being loose. The ESDM has its own curriculum, organized by developmental domains and levels, from which individualized short-term objectives are written. Progress is logged session by session, and there's a formal fidelity check: whether the adult is really applying the model as described, and not a personal version of it.
Therapists, parents, daycare
The model was designed to fit the child's real life, and it exists in more than one format: delivered by trained therapists, taught to parents and caregivers for use in home routines, and adapted for groups in daycare centers and early-childhood classrooms. Because teaching happens within everyday activities, the people around the child can be active participants — not just drop-off and pick-up.
What the evidence shows
Here it's worth separating what's solid from what became marketing. The ESDM is one of the most studied early intervention models, and the picture that emerges is positive, though more modest than it's usually advertised.
- The original study (2010). A randomized trial with 48 children aged 18 to 30 months compared the ESDM — about 20 hours a week, for two years, with therapists and parent training — to a group that received evaluation and referral to community services. The ESDM group gained, on average, 17.6 points on a standardized IQ score, versus 7.0 for the comparison group, with improvement also seen in adaptive behavior. It was the first randomized trial of a comprehensive developmental-behavioral intervention for children this young.
- The replication (2019). A multisite study with 118 children aged 14 to 24 months, across three universities, over 27 months, found an advantage for the ESDM on the primary outcome — receptive and expressive language — but found no significant difference in developmental quotient, autism symptom severity, or adaptive behavior. A partial replication, then.
- The meta-analysis (2020). Pooling 12 studies and 640 children, significant effects appeared in cognition and language, both of moderate magnitude (g ≈ 0.41). There was no significant effect on autism symptoms, adaptive behavior, social communication, or restricted and repetitive behaviors.
An honest reading of the whole pictureThe ESDM helps — mainly with language and cognitive skills, with a moderate effect. There is no consistent evidence that it reduces core autism symptoms or improves adaptive behavior, and the broad cognitive-score gains from the 2010 study were not repeated in the larger replication. This doesn't disqualify the model: it places it in the realistic spot of one good option among others with empirical support, not a transformative treatment.
"How many hours a week?"
This is the question that worries families most, and the research's answer is surprising. A multisite randomized trial with children aged 12 to 30 months compared, at the same time, two intensities — 15 and 25 hours a week, over 12 months — and two styles of evidence-based early intervention — the ESDM itself and early intensive behavioral intervention (EIBI). It found no significant difference between either the styles or the intensities, in any of the four outcomes assessed: autism severity, receptive language, expressive language, and nonverbal ability.
In other words: the idea that "more hours is always better" doesn't automatically hold up. This matters because the race for high hourly loads costs families dearly, takes up the child's entire childhood, and — according to this data — doesn't guarantee a better outcome.
What the model doesn't promise
No early intervention program "takes a child off the spectrum," and the ESDM is no exception. The goal is to broaden communication, learning, and participation — not to manufacture a child who appears non-autistic. Be wary of any service that:
- promises recovery, a "cure," or exit from the diagnosis;
- guarantees a specific outcome in exchange for a number of hours;
- treats autistic traits that cause no harm — such as stereotypies — as targets to eliminate just because they look different;
- shows no data, no written objectives, and no generalization plan for life outside the session.
Watch out for the "window" argumentStarting early is good and has support — but "it's now or never" is often used to sell expensive packages to frightened families. Development continues past early childhood, and no child loses the chance to learn by starting later. Decisions about intervention should be made with information, not fear.
Test your understanding
Quick quiz
Three questions to lock in the concept. You'll see the explanation right away.
Further reading
- Rogers, S. J. & Dawson, G. Early Start Denver Model for Young Children with Autism: Promoting Language, Learning, and Engagement. Guilford Press, 2010.
- Dawson, G. et al. "Randomized, Controlled Trial of an Intervention for Toddlers With Autism: The Early Start Denver Model". Pediatrics, 125(1):e17–e23, 2010.
- Rogers, S. J. et al. "A Multisite Randomized Controlled Two-Phase Trial of the Early Start Denver Model Compared to Treatment as Usual". Journal of the American Academy of Child & Adolescent Psychiatry, 2019.
- Fuller, E. A., Oliver, K. et al. "The Effects of the Early Start Denver Model for Children with Autism Spectrum Disorder: A Meta-Analysis". Brain Sciences, 10(6):368, 2020.
- Rogers, S. J. et al. "A Multisite Randomized Controlled Trial Comparing the Effects of Intervention Intensity and Intervention Style on Outcomes for Young Children With Autism". Journal of the American Academy of Child & Adolescent Psychiatry, 2021.
This entry is educational and does not replace assessment, diagnosis or treatment by a qualified health professional. Every person is unique — for a specific case, seek individual guidance.