Abstract:A one‐mg oral overnight dexamethasone suppression test (DST) was conducted on 22 inpatients with eating disorders. To confirm that the dexamethasone tablets had been ingested, we measumd the plasma concentrations of dexamethasone the next morning (at 0900 hr after DST). The diagnosis of anorexia nervosa and bulimia was made according to the criteria for DSM‐III, respectively. Of the 22 patients with eating disorders, 16 satisfied the criteria for anorexia nervosa and 6 for bulimia. The DST was carried out within 2 weeks of hospitalization on each patient. The subjects were given 1 mg of dexamethasone in the evening (at 2300 hr) and blood samples were collected the following day (at 0900, 1600 and 2100 hr, respectively). The plasma cortisol and dexamethasone levels were concurrently determined by RIA. The criterion for non‐suppression was a failure to suppress the plasma cortisol levels below 5.0 μg/dl in any one of the three samples. All but one patient with bulimia had ingested the dexamethasone. Thirteen (62%) of 21 patients with eating disorders were nonsuppressors. We found a significant positive correlation between the plasma cortisol levels at 1600 hr or 2100 hr and a decrease in ideal body weight (n = 16, r = 0.613, p<0.05; r = 0.75, p<0.01, respectively) and a significant inverse relationship between the plasma dexamethasone levels at 0900 hr and the plasma cortisol levels at 1600 hr was recognized (n = 21, r ‐ 0.631, p<0.01). These results suggest that the blood dexamethasone levels as well as body weight might contribute to the abnormalities of DST seen in patients with eating d