Serum albumin concentration is commonly used as an index of nutritional status and as an indicator of nutritional response in hospitalized patients receiving total parenteral nutrition (TPN). One hundred thirty-nine cancer patients receiving TPN for at least two weeks were studied. Albumin intake, serum albumin, fluid balance, and weight change was monitored from 14 to 100 days of TPN. Patients were classified into three groups: A) patients receiving no exogenous albumin; B) patients receiving less than 25 grams of exogenous albumin; and C) patients receiving at least 25 grams of exogenous albumin during their course of TPN. Linear regression analysis of serum albumin levels vs. time on TPN showed a minimal positive correlation for patients in groups B and C (r = 0.154 and r = 0.183, respectively). Further analysis showed a significant elevation of serum albumin levels only in patients in group C (p less than or equal to 0.05). Contingency table analysis showed statistically significant increase in the incidence of sepsis in patients treated with exogenous albumin (X2 = 10.50, df = 2, p less than 0.01). There was no relationship between the change in serum albumin concentrations and the number of patient deaths. In addition, no relationship between tumor burden and subsequent response of serum albumin levels were identified. Serum albumin levels do not increase in cancer patients receiving TPN, unless exogenous albumin is given. Serum albumin appears to be a poor index of nutritional response in cancer patients receiving TPN.