One purpose of this report is to illustrate that calculating the rate of excretion of osmoles in the urine can be of value in the differential diagnosis of hypernatremia and polyuria. A second purpose is to illustrate a clinical example where the osmolality of the urine did not reflect the lack of action of ADH. A patient with ethanol intoxication seemed to have central diabetes insipidus on clinical grounds. However, the osmolality of the urine was 287 mosm/kg H2O, a value which made this diagnosis unlikely. Since the concentration of ethanol in plasma was 119 mmol/L, we suspected that the urine contained an appreciable quantity of alcohol; this might obscure the lack of action of ADH. A study was performed to document the quantitative relationship between the concentrations of ethanol in plasma and urine. The concentration of ethanol in the urine was approximately 1.4-fold greater than in plasma. Using this correction factor, the osmolality of the urine adjusted for ethanol in the patient was only 120 mosm/kg H2O, a value more consistent with the diagnosis of central diabetes insipidus.