The diagnosis of refractory ascites in cirrhotic patients carries a poor prognosis and liver transplantation should always be considered in this situation. Identification of patients who will not respond to diuretic therapy usually requires several weeks of observation during which a trial of diuretics is instituted using stepwise increases in dosage in order to classify ascites as refractory. In the present study we evaluated the effect of a single dose of 80 mg intravenous furosemide on urinary sodium excretion over 8 hours in cirrhotic patients with ascites responsive to diuretic treatment (group 1; n = 14) and patients with refractory ascites (group 2; n = 15). The test was performed after 3 days without diuretics and patients were on a 80 mEq sodium/day diet. Refractory ascites was defined by the absence of response after 3 months of high doses of diuretics (spironolactone 200 mg/d + furosemide 80 mg/d + metolazone 2.5 mg/d) and the need for repeated paracentesis. The two groups had similar degrees of liver and renal dysfunction as assessed by the Pugh score and creatinine clearance. The effects of furosemide on 8-hour natriuresis was much higher in patients with responsive ascites as compared with patients with refractory ascites (125 +/- 46 vs. 30 +/- 16 mEq; mean +/- SD; P <.0001). A natriuresis lower than 50 mEq/8 hours was observed in all group-2 patients as compared with none from group 1. The present study shows that patients with refractory ascites can be identified quickly and accurately by using this simple furosemide-induced natriuresis test, which could be very useful to select patients for liver transplantation.