Seventy-two patients undergoing routine surgical procedures under propofol-remifentanil anaesthesia were randomly assigned to receive either standard clinical practice (n = 35) or standard practice plus monitoring of depth of anaesthesia with M-Entropy (n = 37). Patients in the standard practice group received more propofol than the entropy group (mean (SD) 95 (14) vs 81 (22) microg.kg(-1).min(-1), respectively; p < 0.01), and less remifentanil (0.39 (0.08) vs 0.46 (0.08) microg.kg(-1).min(-1), respectively; p < 0.001). Loss of consciousness was best predicted by BIS (prediction probability (P(K)) 0.96) and response entropy (P(K) 0.93), whereas emergence was best predicted by response entropy (P(K) 0.94). The frequency of unwanted patient responses was higher in the standard practice group than in the entropy group (47 vs 27 total events, respectively; p < 0.01). Both regimens resulted in fast recovery with no clinical advantage for either one. There were no significant differences in haemodynamic parameters, postoperative nausea and vomiting or satisfaction with the procedure.