Until quite recently, the only stressor considered consistent with the diagnosis of PTSD was a catastrophic, out of the ordinary, trauma that almost anyone could be expected to have a severe reaction to. Thus, PTSD was considered relatively rare among non-military populations. More recently, epidemiologic surveys have suggested that PTSD may be much more prevalent than heretofore recognized, and the DSM-IV has opened the door to a much larger variety of stressors (the "A" criterion). Yet, bereavement is not considered the type of stressor capable of producing PTSD. In this study, 350 newly bereaved widows and widowers were assessed for the prevalence of PTSD, its chronicity, comorbidity, and consequences. The diagnosis of PTSD was made on the basis of questionnaire items approximating the DSM-IV criteria for PTSD. At 2 months after the spouse's death, 10% of those whose spouses died after a chronic illness met criteria for PTSD, 9% of those whose spouses died unexpectedly met criteria, and 36% of those whose spouses died from "unnatural" causes (suicide or accident) had PTSD. Symptoms tended to be chronic in at least 40% of the subjects, almost always were associated with comorbid depression, and created substantial morbidity. The results suggested that PTSD may occur after bereavement, and, by extension, other stressors not recognized by official diagnostic systems. The "A" criterion needs further examination.