Data from the Psychoeducational Profile-Revised (PEP-R) were analysed in a sample of 46 children, aged from 1.7 to 5.11 years, of whom 21 had autistic disorder (AD) and 25 had pervasive developmental disorder not otherwise specified (PDD-NOS). Analysis with a t-test for independent samples revealed a significant difference (p less than 0.05) between children with AD and those with PDD-NOS on both developmental and behavioural PEP-R scales, supporting the utility of the PEP-R in discriminating between two diagnostic groups. This study emphasizes the effectiveness of the PEP-R as a tool for the early assessment of children with pervasive developmental disorders. (Contains 2 figures and 2 tables.)