Because hyponatremia is frequently associated with preceding diuretic treatment and unrestricted fluid intake--conditions which have not been addressed sufficiently in published literature--we studied the pathophysiology and the clinical setting of such hyponatremia in a large group of internal medicine patients. We observed: a) Of an initial 310 patients with chemical hyponatremia only 204 (64%) had an associated plasma hypoosmolality. Since a normal plasma osmolality excludes a disturbance of water metabolism only the 204 patients with hypoosmolar hyponatremia were included in the study. This data shows that plasma osmolality is an essential measurement in any evaluation of hyponatremia. b) In 204 consecutive patients with hypoosmolar hyponatremia the electrolyte disturbance was related to advanced congestive cardiac failure in 25%, decompensated liver cirrhosis in 18%, volume contraction in 28%, syndrome of inappropriate antidiuretic hormone secretion in 19% and renal insufficiency in 4%. c) Plasma vasopressin was measurable in 90% of the 204 patients. It is known that radioimmunoassays to measure vasopressin fail to reliably detect low concentrations of circulating vasopressin (less than 0.5 pg/ml). It may therefore be stated that hypoosmolar hyponatremia was generally characterized by a failure of antidiuretic hormone suppression. d) Mean daily fluid intake of hyponatremic patients was 2.35 +/- 0.15 l. In the presence of stimulated vasopressin this large a fluid intake is bound to worsen the severity of hyponatremia. e) Of 204 patients 126 were treated with diuretics at the time of study. In these patients hyponatremia worsened during such treatments and was associated with evidence of prerenal azotemia. However there were no significant differences between diuretic-treated and -untreated patients with respect to plasma vasopressin stimulation and amount of fluid intake.(ABSTRACT TRUNCATED AT 250 WORDS)