SummaryWe performed prospective and retrospective studies of 96 consecutive patients with acquired immune deficiency syndrome (AIDS) or AIDS-related complex (ARC) to determine the incidence, pathogenesis, and clinical significance of hyponatremia, defined as serum sodium levels ≤130 mmol/L on more than one occasion. Thirty (31.3%), six with ARC and 24 with AIDS, had hyponatremia, and it developed in 20 as outpatients. Age, gender, duration of illness, and weight loss did not differ between groups. The hyponatremic patients had more opportunistic illnesses, includingPneumocystis cariniipneumonia and cytomegalovirus infections, and had a mortality of 70% as compared to 36.4% of the patients without hyponatremia. The probability of 50% survival after diagnosis of human immunodeficiency virus (HIV) infection in the hyponatremic group was 11.5 months, as compared to 39 months for those without hyponatremia, p < 0.001. The probability of 50% survival after development of hyponatremia was 4.5 months and the median length of time to development of hyponatremia was 12.5 months after diagnosis of HIV infection. Eighty-eight percent had hypovolemia and 12% normovolemia. Seventeen of 21 with hypovolemia had no evident source of fluid loss. Two had Addison's disease, and 15 had unexpectedly high urine sodium concentration without evidence of renal or adrenal insufficiency. Hyponatremia occurs commonly in ambulatory patients with ARC or AIDS, appears in patients with higher mortality and morbidity, and does not represent a terminal event. Most patients had hypovolemia and unexpectedly high urine sodium concentration, suggesting defective renal sodium conservation.