A randomized evaluation of 44 malnourished patients, wherein 24 were used as controls and 20 received immediate postoperative parenteral hyperalimentation, indicated that five days of nutritional therapy had no notable effect on the morbidity and mortality experienced by the malnourished patients, in comparison to a third, nonmalnourished group of similar patients. Although central venous nutrition was safely administered without complications immediately after cardiac operations, clinical efficacy of this therapy could not be demonstrated. The inability to establish a dose-response relationship, and hence administer the "optimum" amount of nutrients, may have accounted for the negative results reported. Although preoperative malnutrition is associated with a poorer result after cardiac surgery, postoperative repletion of nutrients appears to be ineffective in reversing this relationship.