The mechanism by which endotoxin administration results in hypoglycemia was evaluated by characterizing [3H]dexamethasone binding and phosphoenolpyruvate carboxykinase activity in hepatic cytosol preparations from treated and control mice. Starved mice were given Escherichia coli 0111:B4 endotoxin or saline intraperitoneally on day 3 after bilateral adrenalectomy. [3H]dexamethasone binding was measured by the charcoal method after the incubation of cytosol preparations with [3H]dexamethasone in the presence or absence of unlabeled dexamethasone. Changes in [3H]dexamethasone binding were found to be time and dose dependent in treated mice. When mice given different doses of endotoxin reached the same stage of morbidity, as indicated by the average time of death, significantly lower glucocorticoid binding was measured. Scatchard analysis of binding isotherms defined a. single class of binding sites. Association and dissociation rate constants and the equilibrium dissociation constant (Kd) were not altered, but the maximum number of binding sites was depressed by endotoxin. The rank order of potency of competitors for [3H]dexamethasone binding, dexamethasone, > hydrocortisone = corticosterone > deoxycorticosterone > progesterone > testosterone = estradiol, was consistent with a glucocorticoid receptor, although the competition was not altered by endotoxin. Endotoxin treatment prevented the glucocorticoid-induced increase in hepatic phosphoenolpyruvate carboxykinase activity. We conclude that the hypoglycemia of endotoxin poisoning is effected, in part, by the inhibition of the glucocorticoid-mediated induction of phosphoenolpyruvate carboxykinase via the down regulation of hepatic glucocorticoid receptors.