We performed a quantitative systematic review of randomised, controlled trials that compared remifentanil to short-acting opioids (fentanyl, alfentanil, or sufentanil) for general anaesthesia. Eighty-five trials were identified and these included a total of 13 057 patients. Intra-operatively, remifentanil was associated with clinical signs of deeper analgesia and anaesthesia, such as fewer responses to noxious stimuli (relative risk 0.65, 95% CI 0.48-0.87), more frequent episodes of bradycardia (1.46, 1.04-2.05), more hypotension (1.68, 1.36-2.07) and less hypertension (0.60, 0.46-0.78). Postoperatively, remifentanil was associated with faster recovery (difference in extubation time of -2.03, 9.5% CI, -2.92 to -1.14 min), more frequent postoperative analgesic requirements (1.36, 1.21-1.53) and fewer respiratory events requiring naloxone (0.25, 0.14-0.47). Remifentanil had no overall impact on postoperative nausea (1.03, 0.97-1.09) or vomiting (1.06, 0.96-1.17), but was associated with twice as much shivering (2.15, 1.73-2.69). Remifentanil does not seem to offer any advantage for lengthy, major interventions, but may be useful for selected patients, e.g. when postoperative respiratory depression is a concern.