OBJECTIVE: The authors examined the longitudinal changes in posttraumatic stress disorder (PTSD) symptom levels and prevalence rates over a 4-year time period among American former prisoners of war (POWs) from World War II and the Korean War. Retrospective symptom reports by World War II POWs dating back to shortly after repatriation were examined for 1) additional evidence of changing PTSD symptom levels and 2) evidence of PTSD cases with a long-delayed onset. METHOD: PTSD prevalence rates and symptom levels were measured by the Mississippi Scale for Combat-Related Posttraumatic Stress Disorder. For the longitudinal portion of the study, participants were 177 community-dwelling World War II and Korean POWs. For the retrospective portion, participants were 244 community-dwelling World War II POWs. RESULTS: PTSD prevalence rates and symptom levels increased significantly over the 4-year measurement interval. Retrospective symptom reports indicated that symptoms were highest shortly after the war, declined for several decades, and increased within the past two decades. Long-delayed onset of PTSD symptoms was rare. Demographic and psychosocial variables were used to characterize participants whose symptoms increased over 4 years and differentiate participants who reported a long-delayed symptom onset. CONCLUSIONS: Both longitudinal and retrospective data support a PTSD symptom pattern of immediate onset and gradual decline, followed by increasing PTSD symptom levels among older survivors of remote trauma.