A randomized controlled design was employed to evaluate a social skills intervention for children with pervasive developmental disorders. Aims included evaluating the acceptability of the program and gathering preliminary evidence on efficacy. Forty-four children, ages 8-11 years, were randomly assigned to treatment or wait list. Treatment consisted of a 16-week group intervention designed to teach appropriate social behavior. Between group comparisons showed that children in treatment were rated as improved on the primary outcome measure, (unblinded parent report), but not on the secondary outcome measure, a parent questionnaire. Parents reported a high level of satisfaction with the intervention. The study supports the feasibility of this intervention to families and highlights challenges for future research in social skills intervention.