The requirement for educational provision for all children to be evidence based has been highlighted in many government policies.¹ The significant additional costs of educating children with autism, and difficulties in recruiting adequately trained teachers, have placed increasing pressure on education authorities to provide more cost-effective instruction methods for pupils with autism. Computer-assisted interventions (CAI) have been suggested as a possible means of meeting the educational needs of this group, and findings from mostly small, uncontrolled studies have indicated positive improvements in areas such as academic ability, social and communication skills, and adaptive behavior.² The present study by Pellecchia et al.³ describes an evaluation of one such program: TeachTown: Basics. This is a computer-assisted intervention that aims to increase students’ vocabulary, listening skills, socio-emotional development, and academic and cognitive abilities. It is designed for children with autism or other developmental disabilities with a nonverbal mental age equivalent of 2 to 7 years. A brief (20 minutes per day for 3 months), small-scale (n = 22, children 3−6-years of age), randomized control trial (RCT) of TeachTown: Basics, previously conducted in the Los Angeles Unified School District,⁴ produced some positive but generally nonsignificant changes in language and cognitive measures compared with a control group. Although the authors themselves were cautious about the results, this very limited evidence from a single RCT was enough for the School District of Philadelphia (the eighth largest district in the country) to decide to roll out the program to students in all of its kindergarten to second grade autism support classrooms.