Large volume paracentesis, while effectively relieving symptoms in patients with tense ascites, has been generally avoided due to reports of complications attributed to an acute reduction in intravascular volume. Measurements of plasma volume in these subjects have been by indirect methods and have not uniformly confirmed hypovolemia. We have prospectively evaluated 18 patients (20 paracenteses) with tense ascites and peripheral edema due to chronic liver disease undergoing 5 liter paracentesis for relief of symptoms. Plasma volume pre-and postparacentesis was assessed by a "'I-labeled human serum albumin dilution technique as well as by the change in hematocrit and postural blood pressure difference. No significant change in serum sodium, urea nitrogen, hematocrit or postural systolic blood pressure difference was noted at 24 or 4 8 hr after paracentesis. Serum creatinine at 24 hr after paracentesis was unchanged but a small but statistically significant increase in serum creatinine (+0.1 f 0.3 mg per dl, p < 0.01) was noted at 48 hr postparacentesis. Plasma volume changed -2.7% (n = 6, not statistically significant) during the first 24 hr and -2.8% (n = 12, not statistically significant) during the 0-to 48-hr period. No complications from paracentesis were noted. These results suggest that 5 liter paracentesis for relief of symptoms is safe in patients with tense ascites and peripheral edema from chronic liver disease.