1. Healthy male volunteers underwent 10 days of hospitalized protein-calorie starvation and a subsequent 10 day repletion phase with complete intravenous nutritional support (IVF). Non-protein calories were provided as either all D-glucose or as 50% D-glucose/50% lipid. 2. In comparison with starvation, whole-body protein breakdown, as assessed by [15N]glycine, [13C]leucine and urinary excretion of 3-methylhistidine (3-MH), was diminished during IVF. The administration of parenteral nutrition did not specifically suppress peripheral tissue protein breakdown, as measured by extremity 3-MH efflux. 3. Despite the differential insulin response to D-glucose/amino acid (50 +/- 6 m-units/ml) as compared with the D-glucose/lipid/amino acid regimen (25 +/- 4 m-units/ml), there was no difference in nitrogen retention between the regimens. Indirect calorimetric determinations revealed that oxidation of substrate during IVF was related to the proportion of D-glucose and lipid infusion.