One hundred and sixty patients with a primary diagnosis of obsessive-compulsive disorder were enrolled in a multicentre, randomized, double-blind, placebo-controlled study of fluvoxamine. After a placebo washout phase, patients were randomized to treatment with placebo or fluvoxamine (100-300 mg/day) for 10 weeks. Seventy-eight patients in each group were evaluable for efficacy. Fluvoxamine was significantly more effective than placebo as assessed by the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS), the National Institute of Mental Health Obsessive-Compulsive (NIMH-OC) scale and the Global Improvement item of the Clinical Global Impression (CGI) scale. The percentage of patients classified as "responders" (much or very much improved according to the Global Improvement item) was also significantly higher in the fluvoxamine group from Week 6 onwards, with 33.3% of fluvoxamine-treated patients and 9.0% of those given placebo classified as "responders" at endpoint. The "responders" to fluvoxamine experienced a substantial clinical benefit as reflected in decreases in their Y-BOCS and NIMH-OC scores. Fluvoxamine was well tolerated with the majority of adverse events considered mild or moderate.