Standardization of a therapeutic approach for ascites due to chronic liver disease was prospectively tried in 100 patients. The four progressive steps were: I—relative bed rest associated to with-drawal of toxic agents, and restriction of salt (40 mEq) and water (1.000ml per day) for 5 to 7 days. In twenty two patients a good diuretic response was observed. Step II, initiated only for patients who did not respond to step I, corresponded to the introduction of espironolactone, 150 mg/day, for at least 6 days. A good response was defined as an increased diuresis and weight loss between 300 and 900g/day. 57% of the patients included in step II, showed such a good diuretic response. In step III the dosage of espironolactone was increased to 300 mg/d and 56% of the patients included in this step presented good diuretic response. Out of 100 patients who began treatment only 12 needed a IV step treatment which consisted of the association with furosemide in the dosage of 40 to 80mg/d. Apart from two patients who did not respond, developing spontaneous bacterial peritonitis during treatment and dying, the remaining 10 cases had a good diuretic response. General mortality due to chronic liver disease occurred in 10% of the patients during treatment for ascitis. Hepatic encephalopathy developed in 22 cases but in only four it was due to the diuretic therapy and reversed. Hyperpotassemia in 16 cases and hyperazotemia in 24 cases were generally transient.