Previous studies have suggested that treatment of ascites in cirrhotic patients by repeated paracenteses and albumin infusion is fast, effective and safe. In one of these studies including patients with hyponatremia or renal impairment, this treatment was associated with a reduction of duration of hospital stay in comparison with large dose diuretics. The aim of this randomized study was to compare paracentesis with albumin perfusion and low dose diuretics in cirrhotic patients with ascites, but without hyponatremia or renal impairment. Twenty-six patients (group 1) were treated with paracentesis (4 L/day) and 27 patients (group 2) were treated with spironolactone (225 to 300 mg/day), associated with furosemide (40 to 80 mg/day), when inefficient alone. Ascites and peripheric edema disappeared more rapidly in group 1 than in group 2, 8.6 +/- 9.6 vs 13.5 +/- 6.7 days (P = 0.001) and 4.1 +/- 2.6 vs 10.5 +/- 6.5 days (P = 0.001) respectively. During hospitalisation, the incidence of complications was higher in group 2 than in group 1: 56 vs 26% (P = 0.03). Hyponatremia occurred in 30% of patients in group 2 and 4% of patients in group 1 (P = 0.04). The duration of hospital stay was shorter in group 1 (15.0 +/- 10.4 days) than in group 2 (21.0 +/- 11.7 days) (P = 0.007). During follow-up, ascites reappeared in 32% of patients in group 1 and 57% of patients in group 2 (P = 0.09). At 3 months, one patient in group 1 and 2 patients in group 2 developed spontaneous peritonitis whereas survival was similar in both groups.(ABSTRACT TRUNCATED AT 250 WORDS)