In a randomized prospective study the efficacy and side effects of xipamide versus the combination spironolactone/furosemide in the treatment of cirrhotic ascites were studied. Out of 27 patients four responded to a basic treatment consisting of salt and water restriction and one had to be excluded because of deterioration of kidney function. The remaining 22 patients were randomized to additional treatment with either 20 mg xipamide/day (group I) or 200 mg spironolactone/day combined with 40 mg of furosemide every other day (group II). A response to treatment during the first 4 days was seen in 7 of 11 patients of group I versus only 3 of 11 patients in group II. In the latter group 7 of 11 patients finally responded after 8 days of treatment. Responsiveness to either diuretic treatment strongly depended on pretreatment fractional Na excretion, FENa. The resistance to diuretic treatment can be predicted by a FENa less than 0.2%, and could be overcome by additional strategies known to reduce avid proximal Na reabsorption. Xipamide frequently induced hypokalemia, whereas hyperkalemia was seen following treatment with spironolactone/furosemide. Kidney function remained stable during either diuretic treatment.