Although evidence has suggested that social skills interventions yield social and symptom benefits for autistic children, significant variability in outcomes between studies has raised important questions regarding efficacy and moderators of intervention outcomes (i.e., which interventions yield positive effects and which autistic children are most likely to benefit). The efficacy of a comprehensive psychosocial program for autistic children, ages 7-12 years (N = 88) was tested in a prior randomized controlled trial (RCT). Significant effects favoring the program group (versus waitlist controls) were found at posttest on measures of autism-feature severity, social/social-communication skills, social-cognitive understanding (nonliteral language), and global social skills, and these were maintained at 4-6 week follow-up. In this exploratory study, demographic and clinical variables were tested as potential moderators of outcomes from the prior RCT. Moderation effects were not evident for demographics, or child IQ, expressive language, autistic diagnostic symptoms, or baseline co-occurring externalizing or internalizing symptoms. Child receptive language appeared to moderate the outcome of nonliteral language only. Overall, program effects were, with one exception, unrelated to third variables.