It has been hypothesized that the magnitude of proximal sodium reabsorption affects the response to aldosterone antagonists in nonazotemic cirrhotic patients with ascites. To verify this hypothesis, we evaluated intrarenal sodium handling by lithium clearance in 51 nonazotemic ascitic cirrhotic patients and in 23 controls who were maintained on the same low-sodium diet (80 mmol/day). Seven of 51 cirrhotic patients underwent spontaneous diuresis, whereas 44 required diuretic treatment. Treatment was started with spironolactone at a dose of 150 mg once daily. The dose was increased to 300 mg and then to 500 mg once daily if no response ensued. Cirrhotic patients who did not experience ascites mobilization with 500 mg spironolactone were then treated with a combined diuretic regimen that included spironolactone at a fixed dose (500 mg once daily) and furosemide at an initial dose of 50 mg once daily. The dose was increased to 100, 150 and 200 mg once daily if no response was noticed. Response to diuretic treatment was defined as body weight loss greater than 700 gm every 3 days until ascites became clinically undetectable. Nonresponders (43%) to spironolactone showed lower sodium fractional excretion (0.34% +/- 0.28% vs. 0.80% +/- 0.50%; p less than 0.001) because of a lower fractional sodium delivery to the distal tubule (18.2% +/- 5.8% vs. 23.4% +/- 7.2%; p less than 0.025) than responders.(ABSTRACT TRUNCATED AT 250 WORDS)