One in two hundred youths suffers from Obsessive-Compulsive Disorder (OCD), yet relatively few receive appropriate treatment. Both cognitive-behavior therapy (CUT) and medication appear beneficial in controlled studies; however, the relative efficacy of CUT and medication, alone and in combination (COMB), is unknown. Thus, well-designed treatment outcome studies are necessary to improve care for youth with OCD. This revised proposal constitutes one half of a (two-center) Multi-Institutional Collaborative Research Project focused on the treatment of pediatric OCD, resubmitted jointly but separately by Drs. John March at Duke University and Edna Foa at Allegheny University (AU). Using a volunteer sample of 120 (60/site) youth age 8-16 with a DSM-IV diagnosis of OCD, the proposed 5 year treatment outcome study contrasts the degree and durability of improvement obtained across six treatment conditions: 3 active treatments, sertraline alone (SER), OCD-specific cognitive behavior therapy (CUT), both SER and CUT (SER + CUT), and 3 control treatments, pill PBO, pill PBO + Educational Support (ES) and SER + ES. The experimental design covers two phases. Phase I is a 2 (site) x 2 (sertraline or pill PBO) x 3 (CUT, ES or no psychosocial treatment) x 5 (repeated measures) factorial 12 week comparison of SER, CUT, COMB and the control conditions. In Phase II, responders advance to a 16 week discontinuation study to assess treatment durability. The primary outcome measure is the Yale-Brown Obsessive-Compulsive Scale. Assessments blind to treatment status take place at week 0 (pretreatment); weeks 1, 4, 8, 12 (Phase I treatment); and weeks 16, 20, 24, 28 (Phase II discontinuation). Besides addressing comparative efficacy and durability of the specified treatments, we also examine time-action effects, differential effects on specific aspects of OCD, including functional impairment, and predictors of response to treatment.