To investigate the adrenostatic potential of a nonhypnotic low dose etomidate infusion, we administered 0.03 mg/kg etomidate in a bolus injection, followed by constant infusion of 0.3 mg/kg.h for 24 h to 6 patients with severe Cushing's syndrome. The dose-response relationship also was determined in 15 normal subjects. Three groups of 5 received, respectively, doses of 0.03, 0.1, and 0.3 mg/kg.h etomidate for 5 h after an initial bolus dose of 0.03 mg/kg. The response to exogenously administered ACTH [0.25 mg ACTH-(1-24)], injected after the etomidate or control infusion, was determined in all normal subjects. In the six hypercortisolemic patients, serum cortisol concentrations decreased from 1374 +/- 436 nmol/L (mean +/- SEM) to 188 +/- 91 nmol/L after 11 h of etomidate infusion and remained low until the end of the infusion. Cortisol levels returned to pretreatment concentrations by 24 h. Excretion of urinary free cortisol decreased from 1180 +/- 196 to 185 +/- 66 nmol/day. In the normal subjects, administration of etomidate led to a dose-dependent decrease in serum cortisol from about 550 to 83 nmol/L, while 11-deoxycortisol rose from low or undetectable levels up to 346 nmol/L. In response to ACTH, cortisol levels rose in inverse proportion to the etomidate dose. It was, however, significantly reduced compared to normal saline infusion even after the lowest dose. Changes in aldosterone and corticosterone concentrations were similar to those in cortisol, and 11-deoxycorticosterone changed in a pattern similar to that of 11-deoxycortisol. Two of five normal subjects reported tiredness during the highest etomidate infusion. No other side-effects were noted. We conclude that iv administered etomidate in a low nonhypnotic dose reduces serum cortisol concentrations in a dose-dependent manner in both hyper- and eucortisolemic subjects. This study suggests that etomidate at a dose of 0.1 mg/kg.h or lower may be an effective strategy for the control of severe hypercortisolemia.