Although, total paracentesis associated with human albumin substitution has shown to be a rapid, effective and safe treatment of diuretic refractory ascites in advanced liver cirrhosis, it implies high costs and has a limited availability. Therefore an alternative procedure the reinfusion of concentrated ascites has gained popularity in recent years (Smart et al. 1990; Grazioto et al. 1997). It was the aim of the study to compare human albumin substitution vs. reinfusion of ascitic-ultrafiltrate after total paracentesis. 35 patients with cirrhosis and tense ascites received total paracentesis associated with either human albumin (5-8 g/l ascites) (= group A) or reinfusion of an ascitic-ultrafiltrate fluid by means of hemofiltration technique (= group B). The mean volume of ascites removed was 9.41 (2.1-20.0) in group A and 11.41 (6.5-21.0) in group B. No significant differences in serum electrolytes, liver and renal function, coagulation profiles and hormones of the renin-angiotensin-aldosterone system were observed during hospitalization. In both groups sodium excretion increased significantly. 43% of the patients in group B developed pyrexia and chill after reinfusion of the ascitic-ultrafiltrate fluid. In one patient an anaphylactic bronchospasm occurred requiring IUC-treatment. The treatment cost in case of human albumin were 326.-DM vs. 290.-DM for each patient treated with ascitic-ultrafiltrate fluid reinfusion. The probabilities of hospital readmission and survival were similar in both groups during follow-up. The results indicate that i.v. infusion of ascitic-ultrafiltrate fluid is as effective as total paracentesis and albumin infusion in case of diuretic refractory ascites. However, according to the costs of instruments and staff and due to the significant allergic reactions caused by ascitic fluid it cannot be considered as a real alternative to albumin substitution.