The diagnosis of posttraumatic stress disorder (PTSD) rests on several core assumptions, particularly the premise that a distinct class of traumatic events is linked to a distinct clinical syndrome. This core assumption of specific etiology ostensibly distinguishes the PTSD diagnosis from virtually all other psychiatric disorders. Additional attempts to distinguish PTSD from extant conditions have included searches for distinctive markers (e.g., biological and laboratory findings) and hypothesized underlying mechanisms (e.g., fragmentation of traumatic memory). We review the literature on PTSD's core assumptions and various attempts to validate the construct within a nomological network of distinctive correlates. We find that virtually all core assumptions and hypothesized mechanisms lack compelling or consistent empirical support. We consider the implications of these findings for conceptualizing PTSD in the forthcoming edition of the American Psychiatric Association's diagnostic manual.