In previous studies, fluoxetine has been shown to increase light, nonrestorative sleep and to decrease sleep efficiency. This work also suggested that women are more sensitive to the alerting effects of selective serotonin reuptake inhibitors on sleep characteristics. The purpose of the present study was to evaluate the effects of fluoxetine on quantitative sleep EEG. The effects of five and ten weeks of fluoxetine treatment on quantitative sleep EEG were evaluated in 36 patients with major depressive disorders compared to baseline when patients were symptomatic but unmedicated. All patients were treatment responders, achieving symptom remission at week 10. Fluoxetine significantly increased fast-frequency activity and decreased delta during sleep at both 5 weeks and 10 weeks of treatment. Contrary to our expectation, treatment effects on EEG were not stronger in depressed women than men. The protocol also included sleep studies at week 30 of treatment and 7-8 weeks after medication discontinuation. Since only 13/36 (36.1%) patients completed the entire protocol, the effects of fluoxetine on quantitative sleep EEG were compared in study completers versus dropouts. Dropouts showed the largest fluoxetine-induced EEG changes (from baseline to week 10) with greater beta amplitude and lower delta amplitude than study completers. This effect approached significance. It was concluded that fluoxetine is alerting to both standard sleep measure and to quantitative sleep EEG and that these effects may be related to clinical outcome.