Self-efficacy is a potentially useful concept when applied to Social Anxiety Disorder (SAD). The aims of the current study were to examine the psychometric properties of the Self-Efficacy for Social Situations Scale (SESS; Gaudiano and Herbert, 2003) and to investigate the relationship between self-efficacy and anxiety in an adolescent sample with generalized SAD. Results replicated those found in a previous adult SAD sample. The SESS showed high internal consistency and good construct and criterion-related validity. The SESS also predicted subjective anxiety and perceived performance in social role play tests after controlling for social anxiety severity. Furthermore, self-efficacy more clearly predicted self-ratings in contrast to observer ratings of performance in social role play tests. Finally, changes in social self-efficacy were strongly associated with changes in social anxiety symptoms following treatment, even after controlling for changes in fear of negative evaluation. Treatment implications for adolescent SAD are discussed.