Examined the utility of CBT for childhood obsessive-compulsive disorder (OCD) including a preliminary exploration of predictors of response to this form of treatment. A total of 42 youngsters (mean age 11.8 years, 60% female, 52% on medication at baseline) with DSM-IV OCD were treated openly using a developmentally sensitive treatment protocol based on exposure plus response prevention (ERP). The treatment response rate (CGI < 2) was 79% with a mean decrease from baseline in NIMH global scores of 45%. Response was not related to age, gender, baseline medication status, comorbid symptomatology, or therapist experience. Poorer outcome was associated with more severe obsessions and greater OCD-related academic impairment at baseline. When presented in a developmentally appropriate manner, CBT is a useful treatment for childhood OCD. Controlled trials are needed to provide a more rigorous test of this treatment approach and provide better information regarding potential mediators and moderators of outcome.