The diabetic hyperosmolar syndrome occurs predominantly in maturity-onset diabetics and is characterized by extreme hyperglycemia, hyperosmolarity of plasma, and absence of ketoacidosis. A syndrome bearing a marked resemblance to the human disease can be produced in rats. The three prerequisites for the development of the experimental syndrome are: (1) moderate nonketotic diabetes (induced by either streptozotocin or alloxan) ; (2) treatment with hydrocortisone acetate; and (3) water deprivation. Plasma glucose concentrations in cortisol-treated, waterdeprived diabetic rats ranged from 838 to 1,488 mg./100 ml. and were significantly higher than those of control diabetic or ketoacidotic animals. Neither cortisol treatment alone nor water deprivation alone was adequate to raise plasma glucose concentration above that of control diabetic rats. The blood pH, acetoacetate, and β-hydroxybutyrate of cortisoltreated, water-deprived diabetic rats did not differ significantly from the diabetic control values, but all three of these parameters were different from those of ketoacidotic rats. Neither cortisol treatment, nor water deprivation, nor a combination of the two treatments significantly affected these parameters in nondiabetic rats. By analogy with the experimental syndrome, it is suggested that administration of corticosteroids may play a role in the development of the human hyperosmolar syndrome in some patients. Furthermore, the fact that water deprivation was necessary for the production of the experimental syndrome suggests that the extreme dehydration seen in the human hyperosmolar syndrome is not merely a consequence of hyperglycemia, but also a factor in the development of the extreme hyperglycemia.