BACKGROUND: The hypothesis that subcortical disinhibition is the reason for etomidate-induced myoclonus suggest that drugs acting on the subcortical area may reduce myoclonus. To verify the hypothesis, premedication with placebo, etomidate of a small dosage, midazolam and fentanyl were compared. METHODS: Sixty ASA physical status I or II patients undergoing elective surgery were allocated into four groups. All groups were induced with etomidate 0.3 mg/kg and vercuronium 0.1 mg/kg and maintained with 50% N2O and 1.5-2% enflurane. Group I (n = 15) received normal saline 3 ml 5 minutes before the etomidate 0.3 mg/kg administration, group II (n = 15) received 0.05 mg/kg etomidate 50 seconds before the etomidate 0.3 mg/kg administration, group III received midazolam 0.05 mg/kg 5 minutes before the etomidate 0.3 mg/kg and group IV received 2 microgram/kg fentanyl 5 minutes before the etomidate 0.3 mg/kg. In all patients, the grade, starting time, maintenance time of myoclonus and vital signs were checked and compared between the four groups. RESULTS: In group IV, myoclonus did not develope except in one patient and there were no differences in the incidence of myoclonus between the others. All premedicating drugs do not affect vital signs. CONCLUSIONS: We find that fentanyl reduces the incidence of etomidate-induced myoclonus but midazolam and a small dose of etomidate are not effective.