Disorders of thirst and vasopressin secretion present clinically in one of three ways: as hypotonic polyuria (DI), as hypodipsic hyponatremia, and as hyponatremia. In evaluating a patient with DI, the major challenge is to differentiate between primary polydipsia and neurogenic and nephrogenic DI. This is best accomplished through a series of steps that start with simple clinical observation, and progress, as necessary, to more complicated diagnostic procedures (Fig. 1). If the diagnosis is not clear from the clinical setting and the patient's history, the first step is to measure plasma osmolality and sodium under conditions of ad libitum fluid intake. If the results are clearly above the upper limit of normal range, primary polydipsia is excluded and the work-up can proceed directly to administration of vasopressin or DDAVP and/or a measurement of plasma vasopressin levels to differentiate between neurogenic and nephrogenic DI. If basal plasma osmolality and sodium fall within normal range, the standard dehydration test should be performed. If urine osmolality does not increase above that of plasma despite evident dehydration, primary polydipsia is excluded and the effect of vasopressin or DDAVP on urine osmolality should be examined to differentiate between neurogenic and nephrogenic DI. If administration of antidiuretic hormone increases urine osmolality by more than 50 per cent, the patient has severe neurogenic DI. If the increase in urine osmolality is less than 50 per cent, the patient has nephrogenic DI. In patients who do not concentrate urine above that of plasma in response to dehydration, the best approach is to measure plasma vasopressin, osmolality, and sodium after the latter have been increased above normal range by dehydration and/or infusion of hypertonic saline. When these results are plotted on a suitable nomogram (Fig. 2), neurogenic DI can be clearly diagnosed from the relative deficiency of vasopressin. In patients with normal vasopressin levels, primary polydipsia can be differentiated from nephrogenic DI by examining the relationship of urine osmolality to plasma vasopressin (Fig. 3), obtained during dehydration and/or graded vasopressin infusion. In evaluating a patient with sustained hypernatremia, it is only necessary to assess thirst, which can be done by a simple bedside observation. In a patient without obvious neurologic or cognitive impairment, absence of thirst in the face of plasma osmolality above 305 mosm/kg (plasma sodium above 150 mEq/L) is diagnostic for hypodipsic hypernatremia. In a patient who presents with hyponatremia, the main objective is to differentiate between hyper-, hypo-, and euvolemic (SIADH) types