We performed a prospective, randomized study comparing the efficacy and safety of remifentanil, propofol or both for conscious sedation during eye surgery under retrobulbar blockade. Forty-five unpremedicated patients were assigned to receive remifentanil (group R) (n = 15, mean dosage: 0.05 +/- 0.03 microgram kg-1 min-1), propofol (group P) (n = 15, 1.5 +/- 0.5 mg kg-1 h-1) or a combination (group RP) (n = 15, R: 0.03 +/- 0.01 microgram kg-1 min-1; P: 0.7 +/- 0.2 mg kg-1 h-1). Haemodynamic responses were comparable among all groups. Minimum values for respiratory rate were lower in R patients (R: 7 vs. P and RP: 10 breaths min-1). Perioperative blood gas analysis showed differences in maximum carbon dioxide tensions (R: 51.5 vs. P: 48.3 vs. RP: 45.5 mmHg) and decrease in minimum pH values (R: -0.06 vs. P: -0.0 vs. RP: -0.01). All group P patients reported mild to intense pain during retrobulbar block, while 53% of the group R patients were free from pain. In group RP, 60% of patients experienced no pain and the remaining 40% reported mild pain only. Remifentanil, applied as the sole agent, provided superior pain relief and patient comfort when compared with propofol, but produced greater respiratory depression and postoperative nausea. The combination of remifentanil and propofol provided haemodynamic stability, adequate spontaneous respiration and pain relief, with a low risk of untoward side effects.