Approximately one-third of persons with depression do not respond to antidepressant monotherapy. Studies suggest that atypical antipsychotic augmentation may benefit these patients. We investigated the longer-term efficacy of risperidone augmentation of serotonin-selective reuptake inhibitor treatment for resistant depression. In 57 in- and outpatient centers in three countries, we conducted a three-phase study with 4-6 weeks of open-label citalopram monotherapy, 4-6 weeks of open-label risperidone augmentation, and a 24-week double-blind, placebo-controlled discontinuation phase. A total of 489 patients with major depressive disorder and 1-3 documented treatment failures entered the citalopram monotherapy phase (20-60 mg/day). Patients with <50% reduction in HAM-D-17 scores entered the risperidone augmentation phase (0.25-2.0 mg/day). Patients with HAM-D-17< or =7 or CGI-S < or = 2 were randomized to risperidone or placebo augmentation. The primary outcome was time to relapse during the double-blind phase. During citalopram monotherapy, 434 patients had <50% HAM-D-17 reduction; 299 (68.9%) were fully nonresponsive (<25% reduction) and 135 were partially nonresponsive (25-49% reduction). Of the 386 nonresponders who entered the augmentation phase, 243 remitted and 241 entered the double-blind phase. Median time to relapse was 102 days with risperidone augmentation and 85 days with placebo (NS); relapse rates were 53.3 and 54.6%, respectively. In a post hoc analysis of patients fully nonresponsive to citalopram monotherapy, median time to relapse was 97 days with risperidone augmentation and 56 with placebo (p = 0.05); relapse rates were 56.1 and 64.1%, respectively (p < or = 0.05). Open-label risperidone augmentation substantially enhanced response in treatment-resistant patients, but the longer-term benefits of augmentation were not demonstrated in this study.