The authors conducted two studies to address issues of the dimensionality, scale reliability, and psychometric properties of scores on the Reynolds Adolescent Depression Scale-Second Edition (RADS-2; Reynolds, 2002) in samples of adolescent psychiatric inpatients. In Study 1 (N=262), they used bifactor analysis to further evaluate the general and specific components of the RADS-2. In Study 2 (N=196), they used confirmatory factor analysis to evaluate the fit of a 1-factor model, the original 4-factor model, a second-order model, and a bifactor model to a new sample data. In both studies, the total RADS-2 and content-specific subscale scores showed acceptable estimates of reliability (i.e., scale reliability estimates >.80). Estimates of concurrent validity were also examined. Scores of the RADS-2 total and content-specific subscale scores were useful in differentiating between the responses of youth with mood disorder diagnoses and those with other primary psychiatric disorder diagnoses. The authors also conducted correlation analyses to identify potential correlates for the total RADS-2 scale and the proposed subscale scores. © 2010 Wiley Periodicals, Inc. J Clin Psychol 66:1–22, 2010.