Thirty hypertensive patients scheduled for cholecystectomy or hernia repair under general anaesthesia with thiopentone-fentanyl-nitrous oxide-pancuronium were divided into two groups of 15. One group received metoprolol tablets 200 mg in a slow release form, once daily for at least 2 weeks including the morning of surgery. In addition, metoprolol 15 mg was injected i.v. shortly before the induction of anaesthesia. The other group received placebo tablets and saline. Two patients in the treatment group and one patient in the placebo group were subsequently excluded, because of complications during treatment. Metoprolol significantly reduced arterial pressure both during undisturbed anaesthesia, during intubation and after extubation. A similar tendency was observed also during surgery, although it was not quite significant (P = 0.055). However, metoprolol had no effect on variations in systemic vascular resistance. Mean pulmonary arterial pressures during anaesthesia and surgery were significantly greater in the control, than in the metoprolol, group. Central venous pressure (CVP) and pulmonary arterial occlusion pressure (PAOP) increased significantly in both groups in response to the surgical stimulus. There was no significant difference between the groups in PAOP and CVP. One patient in the metoprolol group had marked bradycardia (minimum heart rate 26 beat min-1) after neostigmine and atropine; otherwise metoprolol pretreatment was tolerated well.