Results of evaluations of adrenal function in 11 patients with carcinoid tumors are presented. Nine patients had tumors that made and secreted serotonin resulting in elevated 5-hydroxyindoleacetic acid (5-HIAA), elevated serum serotonin, and the carinoid syndrome; while two patients had tumors that did not make serotonin and that did not cause elevated 5-HIAA excretion or elevated serum serotonin. All of the patients had normal 24-hr 17-hydroxycorticosteroid excretion. In the group of patients with tumors actively secreting serotonin, the correlation between 17-hydroxycorticosteroid and 5-HIAA excretion (r = 0.44) was not significant. Six of these patients pretreated with cyproheptadine (CYPRO), a serotonin antagonist, experienced a 36% mean decrease in 17-hydroxycorticosteroid excretion, a finding that was not present when three of them were treated with triprolidine (TPRO), an antihistamine. Serum cortisol at 8 a.m. was normal in all patients except two whose values were mildly elevated, and these two patients showed evidence of suppression of ACTH secretion secondary to dexamethasone treatment. There was a significant positive correlation between serum-cortisol concentrations and 5-HIAA excretions (r = 0.73, p less than .05). Normal diurnal variation was present in six patients in whom it was determined. The serum-cortisol response to insulin-induced hypoglycemia in six patients who had carcinoid tumors actively secreting serotonin was not statistically different from that of 12 normal volunteers. Comparisons between these two groups were difficult because the carcinoid patients' fall in blood sugar was 50%, whereas that of the control group was to 38% of the fasting glucose concentration. Six patients with actively secreting carcinoid tumors responded to standard metyrapone testing with a mean increment of 22.8 +/- 2.5 mg/day in 17-hydroxycorticosteroids. This response was statistically different from the increment of 13.8 +/- 5.3 mg/day in 17-hydroxycorticosteroid excretion found in 34 age-matched hospitalized control patients. When the tests were repeated in four of the patients with carcinoid tumors after pretreatment with CYPRO, the increment in 17-hydroxycorticosteroid excretion was reduced well below the mean increment of the control group. Peak serum 11-deoxycortisol (Compound S) values during the test were also reduced. This decrease in the metyrapone response after CYPRO pretreatment was not due to changed peripheral cortisol metabolism, altered adrenal responsiveness to ACTH, interference with recovery of 17-hydroxysteroids by the Porter-Silber reaction, altered metyrapone metabolism, or reduced renal clearance of Compound S. These changes in adrenal response to metyrapone were not seen when the patients were pretreated with TPRO. Our data suggest that the alterations in adrenal function in our patients may be related to elevated serum serotonin. If CYPRO acts by antagonizing serotonin, these data may give support to the idea of serotoninergic control of cortisol secretion.