Examples are reported of PTSD cases showing full symptomatology (intrusive imagery, avoidance behaviour, disordered arousal) in the absence of a single, acute, dramatic trauma of the kind required by the current DSM‐III‐R definition. Such trauma is thus not a necessary condition for PTSD, and other evidence shows it to be not a sufficient condition. It is suggested that the DSM‐III‐R Axis IV distinction between acute and enduring psychosocial stressors be incorporated into the definition to distinguish two pathways to stress disorder, post‐traumatic (PTSD) and prolonged duress (PDSD). Differential treatment implications of the two routes are noted.