A validated objective measure of the state of psychological function was used to determine the incidence and course of psychological dysfunction in a group of evacuees from Darwin following disaster caused by a cyclone (Cyclone Tracy). While psychological dysfunction was increased initially (58 per cent) and at ten weeks (41 per cent), it had returned to an Australian general population control level (22 per cent) at 14 months. Factors influencing psychological dysfunction were examined, and it is suggested that the sample faced two different stressors at differing times. Initial psychiatric morbidity was most clearly associated with the experience of thinking that one might die or be seriously injured and therefore conceptualized as a 'mortality stressor'. Psychiatric morbidity at ten weeks appeared to be most closely associated with what has been conceptualized as a 'relocation stressor'. Reasons why psychiatric morbidity decreased to a general population control level are discussed.