Introduction Research highlights the importance of early intervention for children with autism spectrum disorder with better outcomes associated with earlier access to early intensive intervention (EII) programmes. However, there is significant variability in response to EII despite children receiving the same programmes. Methods and analysis A prospective, multisite cohort study using a pre-post design assesses the predictors of early intervention outcomes for children who receive EII through six early intervention services (Autism Specific Early Learning and Care Centres, ASELCCs) across Australia. Child and family characteristics at entry to and exit from ASELCCs are ascertained using measures of autism symptoms (Autism Diagnostic Observation Schedule-2; Social Communication Questionnaire); cognitive, language and developmental skills (Mullen Scale of Early Learning); adaptive function (Vineland Adaptive Behaviour Scale - second Edition); behaviours (Child Behaviour Checklist - 1.5 to 5 years; Restricted Repetitive Behaviour Scale); parental stress (Parent Stress Index-4 Short Form); quality of life (Quality of Life in Autism Scale) and a semistructured family history questionnaire for sociodemographic, family and psychosocial characteristics. Characteristics at entry are used as predictors of outcome at exit following EII approximately 12 months later. The change in score from baseline to exit will be the primary outcome of interest. The mediating role of family and psychosocial factors will also be considered. Ethics approval University of New South Wales Human Research Ethics Committee (HC14267). Dissemination of results Findings will be published in peer-reviewed journals and presented at conferences. A report summarising data and the interpretation of data will be published.