Forty patients with cirrhosis of the liver and tense ascites were randomized to receive either aldactone 400 mg/day and furosemide 80 mg/day (n = 20) or repeated large volume paracentesis (LVP) and infusion of low molecular weight dextran (n = 20). Both treatment groups were similar in clinical and laboratory parameters. Complete mobilization of the ascitic fluid was achieved in all receiving LVP and dextran therapy within 1 week of the treatment, in contrast to the minimal mobilization of the ascitic fluid in patients receiving diuretics even after 2 weeks of therapy. Renal function, the clinical parameters of systemic hemodynamics, serum electrolytes, and hepatic function remained stable in patients receiving LVP and dextran and were similar to those in the diuretic-treated patients. We found no deterioration of these functions in the nonedematous patients treated by LVP and dextran even though the protective effect of edema against LVP was lacking in them. Plasma volume estimation in six nonedematous cirrhotic patients treated by LVP and dextran did not reveal any hypovolemia after complete mobilization of ascites. The frequency of complications and death were similar in the two groups. Dextran infusion is a safe, effective, and low-cost replacement therapy in patients with cirrhotic ascites treated by LVP.