Background: Many patients with obsessive-compulsive disorder (OCD) experience little response to standard treatment with serotonin reuptake inhibitors. Mirtazapine enhances serotonergic function by a mechanism distinct from reuptake inhibition. Because a pilot study suggested effectiveness of mirtazapine in OCD, we conducted a controlled trial. Method: We recruited 30 Subjects, 15 treatment-naive and 15 treatment-experienced, with DSM-IV OCD of >= 1 year's duration and a Yale-Brown Obsessive Compulsive Scale (YBOCS) score of >= 20. In the 12-week, open-label phase, subjects received mirtazapine starting at 30 mg/day and titrated over 2 weeks as tolerated to 60 mg/day. At week 12, responders (YBOCS score decrease > 25%) were randomly assigned, double-blind, to continue mirtazapine or switch to placebo for 8 weeks, including a 1-week, double-blind taper week for placebo subjects. Results: In the open-label phase, the mean +/- SD YBOCS score fell frorn 28.3 +/- 3.7 to 20.3 +/- 8.5 (paired samples t = 4.81, p < .0001). Four Subjects (13.3%) discontinued for side effects. Sixteen subjects (53.3%) (8 treatment-naive, 8 treatment-experienced) were responders and 15 agreed to randomization. Response was independent of comorbid mood disorders. In the 8-week, double-blind, placebo-controlled discontinuation phase, the mirtazapine group's mean YBOCS score fell a mean SD of 2.6 +/- 8.7 points while the placebo group's mean score rose a mean SD of 9.1 +/- 7.5 points (Mann Whitney U = 6.5, p = .005, 1-tailed). All other outcome measures were consistent with mirtazapine's superiority versus placebo. Conclusion: Mirtazapine may be an effective pharmacotherapy for OCD. If our results are replicated, larger double-blind studies would be indicated.