BACKGROUND: Most pharmacotherapy trials in post-traumatic stress disorder (PTSD) have been conducted upon male combat veterans. Outcome studies relating to civilians are therefore needed. AIMS: To demonstrate that fluoxetine is more effective than placebo in treating PTSD. METHOD: Civilians with PTSD (n = 53) were treated for 12 weeks with fluoxetine (up to 60 mg/day) or placebo. Assessments of PTSD severity, disability, stress vulnerability, and high end-state function were obtained. RESULTS: Fluoxetine was more effective than placebo on most measures at week 12, including global improvement (much or very much improved: fluoxetine 85%, placebo 62%, difference 0.24, 95% CI 0.01-0.47; very much improved: fluoxetine 59%, placebo 19%, difference 0.40, 95% CI 0.16-0.64), and high end-state function (fluoxetine 41%, placebo 4%, difference 0.37, 95% CI 0.17-0.57). CONCLUSIONS: Fluoxetine was superior for measures of PTSD severity, disability, stress vulnerability, and high end-state function. The placebo-group response was low when viewed as a broad outcome based on a portfolio of ratings, but was higher with a traditional global rating criterion.