The adequacy of pharmacotherapy for patients with obsessive-compulsive disorder (OCD) in a large, prepaid HMO was studied. An analysis was made of the computerized records for December 1, 1994, through April 30, 1998, for members of a Kaiser Permanente plan in northern California who were six years of age or older and had had continuous membership during an index year (May 1, 1995, to April 30, 1996) (1,728 million members). A total of 880 adults and 168 children and adolescents with chart-review-confirmed OCD and a pharmacy benefit were identified. The percentage of patients with an adequate drug trial, defined as > or = 56 days of continuous treatment with a serotonin-reuptake inhibitor or phenelzine at dosages at or above established minimal effective dosages, was determined. Forty-three percent of the adults and 28% of the children and adolescents who were newly diagnosed with OCD in the index year had an adequate trial of medication in the year after their first visit for OCD. By the second six months after the index year, only 75.2% of newly treated adults and 60.9% of newly treated children and adolescents continuing in the health plan filled at least one anti-OCD prescription. During the second follow-up year, these figures fell to 60.4% and 38.9%, respectively. Continuing-care patients filling a prescription in the index year were more likely than newly diagnosed patients to fill prescriptions in the two-year follow-up period, but their treatment still decreased substantially. Despite the typically chronic course of OCD, many patients with OCD who were enrolled in a large HMO appeared not to receive an adequate trial of pharmacotherapy or ongoing pharmacotherapy.