Normal subjects and patients with antidiuretic hormone (ADH) deficiency were studied to determine the mechanism of the antidiuretic action of clofibrate. Before clofibrate treatment, the patients' ability to concentrate urine with a standardized dehydration procedure correlated with the amount of ADH which was excreted. During clofibrate administration all six patients with ADH deficiency developed an antidiuresis which was like that of ADH, since there was no change in sodium, potassium, total solute, or creatinine excretion. There was a correlation between the patients' ability to concentrate urine during dehydration and the subsequent response to clofibrate, and the excretion of ADH during dehydration correlated with the excretion of ADH on clofibrate therapy. Clofibrate-induced antidiuresis in these patients was partially overcome by ethanol and by water loading. Clofibrate interfered with the ability of patients and subjects to excrete a water load and prevented the water load from inhibiting ADH excretion in the normal subjects. These studies suggested that clofibrate was acting through endogenous ADH and this thesis was supported by the failure of clofibrate to produce an antidiuresis when injected into rats with total ADH deficiency (Brattleboro strain) although an antidiuresis was produced in water-loaded normal rats. When the drug was injected into Brattleboro rats with exogenous ADH, clofibrate either did not alter or it inhibited the action of the ADH. The data demonstrate that clofibrate has a significant ADH-like action. This action appears to be mediated through the release of endogenous ADH.