Objectives. Laparotomy in the elderly population is a frequently performed operation with high mortality. Little is published about the anaesthetic management of these patients. In view of this high mortality, such data would be useful prior to designing studies where the effect of interventions in peri-operative practice might be studied. Design. A postal survey of the proposed management of two hypothetical 75 year olds. Patient A: ASA 2, requiring elective anterior resection. Patient B: ASA 3 requiring emergency laparotomy for bowel obstruction. Setting. The survey was sent to the anaesthetic college tutor in 251 acute National Health Service hospitals in the UK, who was asked to pass it to the anaesthetist working in the emergency theatre on that day. Interventions and measurements. If no reply was received within 6 weeks, the survey was resent. Replies were analysed and the data summarized. Main results. Replies were received from 163 (65%) anaesthetists. Nineteen per cent of emergency lists were supervised by a non-consultant. Pre-operative 'optimization' (including, variously, administration of fluids, oxygen and vasoactive drugs) would have been carried out for patient A by 19% and for patient B by 91%, but this was never 'goal-directed pre-optimization'. Invasive haemodynamic monitoring would have been used by two-thirds for patient A and universally for patient B. Conversely epidural techniques would have been used almost universally for patient A and by 71% for patient B, in whom the use of an epidural would also have been delayed until postoperatively more frequently. For 97%, anaesthetic maintenance would have included a volatile agent, for 33Á45% nitrous oxide and for 10% remifentanil. There was evidence of wide variations in the tolerance of hypovolaemia and hypotension between anaesthetists, but not between patients. The initial management of these complications was generally with fluid administration but subsequent drug management varied considerably. Post-operatively, all respondents would have extubated patient A, half would have returned patient A to the wards and none to intensive care. For patient B, half would have extubated the patient and 87% would have transferred the patient to a critical care bed. Sixty per cent of respondents reported difficulty in accessing critical care beds. Conclusions. Laparotomy in the elderly is managed variably in the UK. This survey identifies considerable variation between anaesthetists in the management of these high-risk patients. It is likely that variations in management impact outcome. Knowledge of current practice has value in planning further rigorous studies of this subject.