Etomidate is known to inhibit adrenocorticosteroid synthesis. The extent and duration of the effects of etomidate (63 +/- 6.4 mg) on spontaneous and stimulated corticosteroid levels, as well as on plasma concentrations of ACTH, beta-endorphin, and catecholamines were examined and compared to those following administration of the new benzodiazepine, midazolam, or of methohexital. Twenty-nine healthy, young, male orthopedic patients were randomized into three groups receiving either etomidate/fentanyl (n = 12), midazolam/fentanyl (n = 8), or methohexital/fentanyl (n = 9). Etomidate caused cortisol levels to decrease from 12.5 +/- 1.2 micrograms/dl preoperatively to 5.9 +/- 0.8 micrograms/dl after operation (P less than 0.001), compared to an increase from 12.0 +/- 1.9 micrograms/dl to 18.5 +/- 2.9 micrograms/dl in the group receiving methohexital. At 6 and 20 h postoperatively, all cortisol levels were normal. The cortisol decrease from 12.5 +/- 1.7 to 7.6 +/- 1.5 caused by midazolam was similar to that following etomidate, but the response to exogenous ACTH was significantly impaired in patients receiving etomidate as compared to those receiving midazolam. ACTH and beta-endorphin levels increased in patients receiving etomidate, presumably as a result of the interruption of negative feedback due to cortisol synthesis inhibition. Midazolam on the other hand prevented the increase of ACTH and beta-endorphin levels. Etomidate completely suppressed spontaneous aldosterone levels (from 33 +/- 6.7 to 7 +/- 2.1 pg/ml), as well as the response to stimulation with exogenous ACTH without affecting serum electrolytes. Etomidate had no influence on plasma catecholamines, but midazolam attenuated the stress-related epinephrine increase.