To determine whether gains from exposure therapy are lasting in patients with chronic obsessive compulsive disorder, the authors followed up 34 (85%) of 40 such patients who had been treated 6 years earlier with exposure therapy for 3 or 6 weeks and with clomipramine or placebo for 36 weeks. Severity of obsessive compulsive disorder was assessed by rating the discomfort caused by the time devoted to four target rituals, the Behavioral Avoidance Test, and the Compulsion Checklist. Mood was assessed by the 17-item Hamilton Rating Scale for Depression, the Wakefield Self-Assessment Depression Inventory, and the Anxiety scale. In addition, the patients' general adjustment was assessed. The authors found that the group as a whole remained significantly improved on obsessive compulsive symptoms, work and social adjustment, and depression; however, the group returned to pretreatment levels (slight to moderate) of general anxiety. They found that neither clomipramine nor placebo affected long-term outcome and that the majority of patients who were taking clomipramine or other antidepressants at follow-up were no more improved that those who were not taking antidepressants. Better long-term outcome correlated with more exposure therapy (6 weeks of therapy vs. 3 weeks) and with better compliance with the exposure therapy homework. The best predictor of long-term outcome was improvement at the end of treatment. Subjects who had initially been most depressed were more likely to receive psychotropic medication during follow-up. Initial severity of illness did not preclude benefit from exposure therapy.