A reduction in effective (nonportal) plasma volume is considered the basis for renal sodium retention, a spontaneous reduction in glomerular filtration rate (GFR), and a fall in GFR occurring during drug-induced diuresis in patients with cirrhosis and ascites. In the present study the concept of a reduced effective plasma volume in cirrhosis is challenged by two lines of evidence, even though effective plasma volume itself could not be measured. (a) Total plasma volume failed to rise in 10 patients with the spontaneous loss of ascites, the appearance of sodium in the urine, and a rise in GFR. Portal pressure remained constant in these patients as ascites left, suggesting that effective plasma volume had not increased while portal plasma volume decreased. (b) Reduction of GFR could not be prevented in five patients with cirrhosis and ascites while total plasma volume was prevented from falling with albumin infusions during drug-induced diuresis. Reduction of GFR during drug-induced diuresis in 15 patients with cirrhosis and ascites was completely reversed with saline infusion despite continued diuresis with the identical drugs, excluding drug nephrotoxicity as the cause for the reduced GFR.The ascites of cirrhosis might no longer be regarded as a cause of effective plasma volume contraction, stimulating renal sodium retention and a reduction in GFR. More likely, this form of ascites is a result of plasma volume expansion and sodium retention. The causes for renal sodium retention and a spontaneous reduction in GFR remain unknown. The cause for a fall in GFR during drug-induced diuresis also remains unknown, but effective plasma volume contraction and drug nephrotoxicity seem excluded.