Eleven Vietnam veterans with war-related posttraumatic stress disorder (PTSD) were randomly assigned to 10 sessions of either virtual reality exposure (VRE) therapy within a computer-generated virtual Vietnam environment or present-centered therapy (PCT) that avoided traumatic content and utilized a problem-solving approach. Participants were assessed at pretreatment, posttreatment, and 6 months posttreatment by an independent assessor blind to treatment condition. Nine participants completed treatment with one dropout per condition. No significant differences emerged between treatments, likely due to insufficient power. Although comparison of mean changes in PTSD symptoms for the VRE and PCT conditions yielded a moderate effect size (d = 0.56) in favor of VRE at 6 months posttreatment, changes in PTSD scores were more variable, and therefore less reliable, within the VRE condition. The utility of VRE with older veterans with PTSD is discussed.