The purpose of this study was to evaluate the factor structure of the Liebowitz Social Anxiety Scale for Children and Adolescents (LSAS-CA). The LSAS-CA was administered to 225 children and adolescents as a component of various clinical studies. In addition, other measures of psychopathology and impairment were administered to a subgroup of the sample. Confirmatory factor analyses of the social interaction and performance subscales for the anxiety and avoidance ratings yielded poor fit indices. Exploratory factor analysis supported a two-factor solution with a higher order factor for the LSAS-CA anxiety and avoidance ratings. Based on item content, factors were named Social and School Performance. The internal consistency of the factors was high and the convergent and divergent validity was supported vis-a-vis correlations with measures of depression and social anxiety, and clinician ratings of impairment and functioning. Findings suggest that the anxiety and avoidance ratings are best explained by a two-factor solution that measures social anxiety and avoidance in social and school performance interactions. This factor structure appears to be a reliable and valid framework for assessing childhood social phobia.