Data are presented on the metabolic and endocrine effects of intravenous infusions in normal fasting man observed under highly controlled conditions over a period of six to eight days duration. There are comparative data on a variety of intravenous feeding programs. The data on total starvation are based on studies from the literature, some of which were carried out in this laboratory. The data on low dose glucose, high dose glucose, glycerol, fat emulsion, and amino acids, each given separately, demonstrate changes seen with simple infusion of a single substrate in fasting. These data are now compared with the utilization of amino acid infusions when accompanied by low dose glucose, high dose glucose, glycerol, and fat emulsion. In all, nine experimental intravenous feeding programs are presented, based on data from 35 subjects observed over a total of 370 subject-days. The findings show a strong interaction between glucose or lipid and protein metabolism. In fasting, glucose had protein sparing effect, most evident when given at high dose. Glycerol, in an amount equal to that contained in 2000 ml of ten per cent fat emulsion, had a mild protein sparing effect. Fat emulsion was no more effective. When amino acids were given alone, normal fasting human subjects were always in negative nitrogen balance with the daily nitrogen loss half that seen in starvation alone. Although amino acids given alone have a protein sparing effect, this is accomplished only at the expense of a high nitrogen excretion including an amount equivalent to the entire infusion plus an additional loss from the body's native proteins. The provision of energy yielding non-protein substrates with the amino acids markedly improved nitrogen economy in the following order: glycerol, low dose glucose, fat emulsion and high dose glucose. When caloric provision with glucose approached the isocaloric level for normal diet, the utilization of amino acids was maximized. When given with amino acids, fat emulsion was more effective than the available glycerol alone. THE ACCOMPANYING ENDOCRINE AND BIOCHEMICAL CHANGES SUGGEST THAT THE MILIEU FOR IDEAL UTILIZATION OF INFUSED AMINO ACIDS IS VARIABLE: ketones at low range (carbohydrate) or moderately elevated (fat emulsion); insulin elevated (carbohydrate) or unchanged (fat emulsion). The utilization of the infused amino acids was markedly improved in both endocrine settings, suggesting that it is the provision of energy as substrate as well as the endocrine setting that determines amino acid utilization. There were other changes in plasma intermediates, particularly fatty acids, glucose and urea, all consistent with the concept that when amino acids are given without other substrates, the amino acids must be maximally utilized for gluconeogenesis. When other substrates are provided (particularly glucose at high dose) then this mandate no longer exists and protein synthesis from the amino acids is favored. Several of the plasma amino acid concentrations responded to glucose when added to amino acid infusion. Amino acids alone produced increases in concentration of all the amino acids found in the infusion with the exception of alanine, arginine, and threonine. Many of these increases were abated by the addition of glucose to the amino acid infusion, suggesting an increased utilization rate. Glycerol and fat emulsion, while modulating increases in the plasma amino acid concentration, did so to a lesser extent than did glucose. This lowering of amino acid concentration was unaccompanied by an increase in urinary excretion. The assumption is therefore made that the provision of the added glucose favors the incorporation of amino acid into protein. There is no evidence from these data to suggest that a rising concentration of ketones in the blood favors amino acid utilization or protein synthesis.