JANEWAY and his associates1 in 1944 administered human albumin intravenously to two patients with hepatic cirrhosis, ascites and hypoalbuminemia and noted temporary improvement with control of ascites. Since that time, conflicting reports have appeared concerning the effects of human albumin in treatment of decompensated hepatic cirrhosis.2 The present study is concerned with the effects of intravenously administered salt-poor human albumin on the clinical courses of 34 patients, critically ill with cirrhosis of the liver, hypoalbuminemia and ascites. The substance was given in amounts sufficient to maintain a normal serum albumin concentration for extended periods. The results indicate that human albumin has a place in the therapy of such patients. METHODS A. Techniques of Study. —All patients were kept in bed until there was evidence of continuing clinical improvement. The dietary regimen was as follows: 140 Gm. of protein, 340 Gm. of carbohydrate and 110 Gm. of fat were