The purpose of the study was to investigate the prevalence of posttraumatic stress disorder (PTSD) after prolonged intensive care unit (ICU) treatment and to study the differences between trauma patients and patients who were admitted with other diagnoses. Survivors of surgical intensive care of at least 30 days' duration underwent follow-up examination at about 35 months after discharge from the ICU. Thirty-seven patients were investigated. Seven patients met full DSM-IV criteria for diagnosis of PTSD. All of them had sustained severe multiple injuries. Patients with trauma were at significantly higher risk for developing PTSD than were non-trauma patients. After prolonged ICU treatment, PTSD seems more likely to occur when the reason for admission to the ICU was severe physical injury. The prevalence of PTSD does not seem to be related to injury severity or duration of ICU treatment.