Glucocorticoid negative feedback is exerted in at least two time domains: fast feedback (within minutes of the feedback signal) and delayed feedback (within hours of the feedback signal). Although delayed feedback is known to inhibit ACTH responses to a variety of stimuli in humans, whether there is fast feedback inhibition of the ACTH responses to such stimuli is not known. The purpose of this study was to evaluate the efficacy of a pharmacological injection of cortisol sodium succinate (CORT) as a rapid inhibitor of the ACTH response to surgery in patients undergoing thoracotomy for myocardial revascularization. Thirty patients were premedicated with diazepam and induced with thiopental sodium. They were assigned to one of four groups: group I, general anesthesia was maintained with enflurane (n = 8); group II, patients were anesthetized as in group I, but received a bolus injection of 500 mg CORT within 5 s of the start of surgery (n = 7); group III, anesthesia was maintained with 50-100 mg fentanyl (FENT; n = 8); and group IV, patients were anesthetized as in group III and given CORT as in group II (n = 7). Surgery induced a large increase in plasma ACTH in group I (no CORT, no FENT); the mean plasma ACTH level was 57 +/- 14 (+/- SE) pmol/L 10 min after the start of surgery, and it peaked at 92 +/- 18 pmol/L 50 min after the start of surgery. Administration of CORT at time zero (group II) resulted in a significant but attenuated ACTH response to surgery both 10 min (36.5 +/- 9.7 pmol/L) and 50 min (42.5 +/- 7.3 pmol/L) after the start of surgery. FENT per se (group III) significantly attenuated the ACTH response to surgery (e.g. plasma ACTH was 13 +/- 5 pmol/L 10 min and 21 +/- 7 pmol/L 50 min after the start of surgery). The combination of CORT and FENT (group IV) eliminated the ACTH response to surgery at all time points. In fact, plasma ACTH levels became undetectable (less than 4.4 pmol/L) from 30-50 min after the start of surgery. We conclude that a pharmacological dose of CORT administered at the time of stimulus introduction significantly attenuated the ACTH response to the stimulus (surgery). FENT not only inhibited the ACTH response to surgery per se, but amplified the effect of CORT, such that ACTH actually declined even during a large surgical stimulus. CORT clearly attenuates the ACTH response to surgery in humans in the fast feedback time domain.