BACKGROUND: Insomnia and major depressive disorder (MDD) can coexist. This study evaluated the effect of adding eszopiclone to fluoxetine. METHODS: Patients who met DSM-IV criteria for both MDD and insomnia (n = 545) received morning fluoxetine and were randomized to nightly eszopiclone 3 mg (ESZ+FLX) or placebo (PBO+FLX) for 8 weeks. Subjective sleep and daytime function were assessed weekly. Depression was assessed with the 17-item Hamilton Rating Scale for Depression (HAM-D-17) and the Clinical Global Impression Improvement (CGI-I) and Severity items (CGI-S). RESULTS: Patients in the ESZ+FLX group had significantly decreased sleep latency, wake time after sleep onset (WASO), increased total sleep time (TST), sleep quality, and depth of sleep at all double-blind time points (all p < .05). Eszopiclone co-therapy also resulted in: significantly greater changes in HAM-D-17 scores at Week 4 (p = .01) with progressive improvement at Week 8 (p = .002); significantly improved CGI-I and CGI-S scores at all time points beyond Week 1 (p < .05); and significantly more responders (59% vs. 48%; p = .009) and remitters (42% vs. 33%; p = .03) at Week 8. Treatment was well tolerated, with similar adverse event and dropout rates. CONCLUSIONS: In this study, eszopiclone/fluoxetine co-therapy was relatively well tolerated and associated with rapid, substantial, and sustained sleep improvement, a faster onset of antidepressant response on the basis of CGI, and a greater magnitude of the antidepressant effect.