The factors controlling ascites formation and reabsorption, as well as the relationship of ascites dynamics to renal sodium retention complicating cirrhosis, are not defined. We measured, using labeled albumin, the ascites albumin clearance rate, the plasma-ascites and ascites-plasma albumin filtration rates, the ascites albumin exit rate, and the plasma-ascites and ascites-plasma albumin transfer rates in seven patients with cirrhosis and ascites. Wedged hepatic vein pressure (WHVP), right atrial pressure (RAP), ascites pressure (AP), and serum and ascites oncotic pressure (SOP, AOP) were used to calculate the net hydrostatic (WHVP - AP), oncotic (SOP - AOP) and hydrostatic-oncotic pressure, or "transfer" pressure, favoring ascites formation [(WHVP - AP) - (SOP - AOP)], and the net hydrostatic pressure favoring ascites reabsorption (AP - RAP). Over 4 hours: the ascites albumin exit rate greater than the plasma-ascites albumin transfer rate greater than the ascites-plasma albumin transfer rate (P less than 0.05), and the ascites albumin clearance rate greater than the plasma-ascites and ascites-plasma albumin filtration rates (P less than 0.05). The ascites-plasma albumin filtration rate was inversely related to ascites volume (r = 0.91, P less than 0.01). Calculating the ascites-plasma albumin transfer rate from the extrapolated ascites-plasma albumin filtration rate at an ascites volume of 0 (0.031 L/hr/m2) produced values similar to the mean plasma-ascites albumin transfer rate (0.20 + 0.11 gm/hr/m2 vs. 0.24 + 0.13 gm/hr/m2, not significant).(ABSTRACT TRUNCATED AT 250 WORDS)