Early postoperative recovery was studied using sedation scoring, measurement of flicker fusion frequency and completion of Trieger test figures in 60 male patients who presented for vasectomy under general anaesthesia as day patients. Anaesthesia was induced in groups 1 and 2 (20 patients each) with mean (SD) doses of 0.16 (0.04) mg/kg or 0.16 (0.03) mg/kg midazolam respectively; group 2 received flumazenil 0.55 (0.19) mg after completion of surgery. The remaining 20 patients (group 3) received propofol 1.50 (0.24) mg/kg. Anaesthesia was maintained with isoflurane vaporized in 33% oxygen and nitrous oxide in all patients. Flumazenil tended to improve tests of recovery after midazolam anaesthesia, but early recovery after propofol anaesthesia was associated with better psychomotor test results and less impairment of mental state as judged by sedation and amnesia scoring.