A prospective audit of emergency theatre use for general surgery has been undertaken. Two month periods were studied before and after the introduction of a fully staffed 24-hour emergency theatre. Data were collected using a proforma documenting the time of the decision to operate, the actual time of the operation and the reason for and duration of any delay. After the introduction of the facility the proportion of procedures performed after midnight fell from 29 cases (21.3 per cent) to 7 (6.3 per cent) (p < 0.05). Emergency operating between 0900-1700 hrs increased from 40 cases (29.4 per cent) to 71 (61.3 per cent) (p < 0.05). There was no significant difference in the causes of delay between the two groups, the commonest being queuing for theatre. However, the length of the delay was significantly reduced. That for an appendicectomy was reduced from a median of 4 hrs 40 mins (range 30 mins-18 hrs 45 mins) to 1 hr 29 mins (0-6 hrs 30 mins) (p < 0.01) and for drainage of abscess from 5 hrs 56 mins (15 mins-20 hrs 30 mins) to 1 hr 51 mins (0-4 hrs 30 mins) (p < 0.01). There was no significant difference in the seniority of the surgeon making the decision to operate. In the first part of the audit we identified problems with regard to delay which were addressed by the introduction of the emergency theatre. The audit cycle has been successfully closed improving the care of general surgical emergencies requiring urgent or emergency operations.