Current guidelines for the treatment of bipolar type II (BP II) major depressive episode (MDE) recommend using either mood stabilizer monotherapy or the combination of a mood stabilizer with a selective serotonin reuptake inhibitor (SSRI). These guidelines are the result of concern over SSRI-induced manic switch episodes. We previously showed that fluoxetine monotherapy may be effective as an initial treatment for BP II and BP NOS MDE with a low manic switch rate. We now present the results of a double-blind, placebo-substitution continuation study of fluoxetine monotherapy in BP II and BP NOS patients who have recovered from their MDE. This was a two-phase study. In study phase I, patients received open-label fluoxetine monotherapy 20 mg daily for up to 8 weeks. Responders with a final 17-item Hamilton Rating Scale for Depression (HAM-D) score < or =9 were enrolled into study phase II which consisted of double-blind, placebo-substitution continuation therapy with fluoxetine 20 mg daily for up to 6 months. Outcome measures included the 17-item HAM-D and Young Mania Rating (YMR) scales. Changes in YMR scores were assessed using generalized estimating equation analysis. Relapse was assessed using Kaplan-Meier survival analysis and Fisher's exact test. In study phase II, 43% of fluoxetine-treated patients and 100% of placebo-treated patients relapsed during continuation therapy (P=0.08). The mean increase in YMR score in study phase II was slightly higher in the fluoxetine-treated patients (3.0+/-1.8) versus placebo-treated patients (0.2+/-0.4) (P=0.01). However, this difference was not clinically meaningful. No hypomanic switch episodes were observed during study phase II. Despite the limited sample size resulting in insufficient power to detect statistical significance in relapse rates or change in YMR scores between treatment conditions, these preliminary data appear to support previous observations demonstrating that initial and continuation fluoxetine monotherapy may be safe and effective for some patients with BP II or BP NOS MDE with a low manic switch rate. Larger-scale studies are needed to confirm these findings.