The effect of anaesthesia on the hyperglycaemic and adrenocortical response induced by surgery was studied in patients undergoing abdominal hysterectomy. The study group was anaesthetized with midazolam and alfentanil using a totally intravenous anaesthetic technique. A reference group received anaesthesia with thiopentone, alfentanil and nitrous oxide. Midazolam 0.42 mg.kg-1 was given as a loading infusion followed by a maintenance infusion of 0.125 mg.kg-1.h-1. Alfentanil was given as a bolus dose of 0.075 mg.kg-1 in both groups, followed by a loading infusion of 0.3 mg.kg-1.h-1 for 15 min and a maintenance infusion of 0.065 mg.kg-1.h-1. Increments of alfentanil were given whenever heart rate or systolic blood pressure exceeded pre-induction values by more than 10%. During anaesthesia mean arterial pressure and heart rate were similar in both groups and there was no difference in alfentanil requirement. An immediate increase in blood glucose concentrations was seen following incision, but maximum concentrations were measured in the early postoperative period. Serum cortisol concentrations decreased after induction of anaesthesia. During surgery they returned to pre-induction values, and in the postoperative period they increased to about twice the pre-induction values. It is concluded that midazolam/alfentanil anaesthesia is as effective as anaesthesia induced by thiopentone, alfentanil and nitrous oxide in suppressing the stress-response to surgery until the postoperative period. No signs of prolonged adrenocortical depression were observed.