The present study is the first controlled study that evaluates the effects of cognitive therapy along the lines of Beck (1976) [Cognitive therapy and the emotional disorder. New York: International University Press] and Salkovskis (1985) [Behaviour Research and Therapy, 23, 571-583] in obsessive compulsive disorder (OCD) and compares these effects with those of self-controlled exposure in vivo with response prevention. Seventy-one patients were randomly assigned to either cognitive therapy or exposure in vivo. In each treatment condition seven patients dropped out. Both treatments consisted of 16 sessions. Cognitive therapy as well as exposure in vivo led to statistically significant improvement. Multivariate significant differences suggesting a superior efficacy of cognitive therapy in comparison to exposure in vivo on the obsessive compulsive measures and on the measures for associated psychopathology. However, no univariate differences were found. Further, in both treatment conditions a considerable percentage of the patients was rated as "recovered". Significantly more patients were rated as "recovered" in the cognitive therapy. The results show that this form of cognitive therapy is an effective treatment for OCD and suggest that cognitive therapy may be even more effective than exposure in vivo.