To determine the prognosis of hyponatremia in an unselected population, we collected clinical and laboratory data and determined the outcome of hospitalization for all hyponatremic patients in a general hospital over a three month period. Of the 78 patients studied, 36 (46%) had CNS symptoms while 42 (54%) were asymptomatic. Using a diagnostic algorithm we classified the patients with CNS symptoms into two groups: those with CNS symptoms due to hyponatremia (11 patients) and those with CNS symptoms caused by other factors (25 patients). Twenty-one patients (27%) died during the study. The highest mortality (64%) was in patients with CNS symptoms related to factors other than hyponatremia. Patients with CNS symptoms due to hyponatremia had a mortality rate (9%) similar to that of patients without CNS symptoms (10%). These findings suggest that the relationship between hyponatremia and outcome is probably not causal. Rather, hyponatremia appears to be a marker for severe underlying disease that carries a poor prognosis.