Forty healthy patients undergoing orthopaedic surgery were randomly allocated to receive an initial blood propofol target concentration of either 4 μg.ml−1 or 6 μg.ml−1 for induction of anaesthesia with a 'Diprifusor' target controlled infusion system for propofol, and analgesic supplementation with either nitrous oxide 67% in oxygen or alfentanil 15-20 μg.kg−1.h−1. Anaesthesia was induced within 3 min in 80% and 95% of patients with propofol target concentrations of 4 μg.ml−1 and 6 μg.ml−1, respectively. The frequency of discomfort on infusion was similar for both target concentrations. During maintenance, supplementary doses of alfentanil were required to provide adequate surgical conditions in approximately half of the patients receiving nitrous oxide. There was no statistically significant difference between the target concentration [mean (SD)] of propofol for total intravenous anaesthesia [5.1 (2.0) μg.ml−1] compared with a technique using nitrous oxide [4.6 (1.2) μg.ml−1] supplemented as needed with small doses of alfentanil.