There is growing interest in measuring social disability as a core element of autism spectrum disorders in medication trials. We conducted a secondary analysis on the Aberrant Behavior Checklist Social Withdrawal subscale using data from two federally-funded, multi-site, randomized trials with risperidone. Study 1 included 52 subjects assigned to placebo and 49 subjects to risperidone under double-blind conditions. Study 2 included 49 subjects assigned to risperidone only and 75 subjects assigned to risperidone plus parent training. After 8 weeks of treatment, all active treatments were superior to placebo (effect sizes ranging from 0.42 to 0.65). The findings suggest that the Social Withdrawal subscale may be a useful measure of social disability in acute treatment trials.