This randomized, controlled trial, comparing the Comprehensive Autism Program (CAP) and business as usual programs, studied outcomes for 3-5 year old students with autism spectrum disorder (ASD). Participants included 84 teachers and 302 students with ASD and their parents. CAP utilized specialized curricula and training components to implement specific evidence-based practices both at school and home. A comprehensive set of outcome areas was studied. Hierarchical linear modeling was used to estimate the treatment impact. CAP had small positive impacts on the students' receptive language (effect size of 0.13) and on their social skills as rated by teachers (effect size of 0.19). Treatment effects were moderated by severity of ASD.