The efficacy and safety of fluoxetine (20-80 mg) was compared with placebo in 144 veterans [36.2 years], diagnosed with combat-related post-traumatic stress disorder (PTSD) selected from a 12-week acute and 24-week relapse prevention PTSD trial. In the acute phase, improvements were greater with fluoxetine than placebo in the disease-specific outcome measures: Treatment Outcome PTSD (TOP-8) total scores (SE):-9.05 (0.90) and -5.20 (1.23), p = 0.001; Clinician Administered PTSD Scale (CAPS) total scores:-31.12 (2.72) and -16.07 (4.24), p < 0.001; all CAPS subscores; Davidson Trauma Scale (DTS) total scores; and other general outcome measures. In the maintenance phase, fluoxetine was superior to placebo in sustaining improvement in TOP-8 [-1.01 (0.91) and 1.56 (0.95)] and CAPS [-4.93 (3.54) and 5.48 (3.66)]. The risk of relapse in the placebo arm was significantly greater than in the fluoxetine arm (log-rank test chi 2 = 4.090, df = 1, p = 0.048). Fluoxetine was well tolerated at a mean daily dose of 65 mg.