The formation of ascites in patients with cirrhosis of the liver probably depends on at least three factors: (1) hypoalbuminemia, (2) portal hypertension, and (3) increased amounts of circulating antidiuretic substance. The relative importance of these factors varies with the individual patients. Although the formation and disappearance of ascites do not correlate perfectly with levels of serum albumin in all patients, in the majority of cases the two findings are coincident and apparently related (1, 2). The use of concentrated human serum albumin in these patients for its osmotic effect would be expected to cause a diuresis. The extent to which this diuresis is produced would depend in large part upon the relative importance of the various factors of ascites production, being most marked in those patients in whom hypoalbuminemia is most specifically at fault.The observations of Janeway (3) and Thorn (4) and their co-workers in patients with liver disease demonstrated that the albumin level of the serum could be raised to normal by repeated injections of albumin. Certain of their patients showed a loss of edema and ascites but the oncotic effects were disappointing in their experience.