Body N balance, 3-methylhistidine (MEH) excretion, amino acid (AA) plasma concentration, and fluxes across the leg were investigated both during fasting and during parenteral nutrition of injured patients in order to better understand protein-sparing mechanisms induced by metabolic support in the whole body and in skeletal muscle. Patients were randomized to receive 15 or 30 kcal/kg.day coupled with 0.30 g of N either with standard or branch-chain (BC)-enriched AA solutions. During fasting, patients were highly catabolic (N balance -14.7 +/- 1.2 g N/m2.day, MEH excretion 422 +/- 25 mumol/m2.day) and showed a high efflux of AA N from the leg (5.08 +/- 2.1 g N/m2.day) without difference between the groups. During treatment, body N balance (-5.55 +/- 0.88, p less than .001) and MEH excretion (284 +/- 20, p less than .001) were significantly reduced without difference among the groups; also, AA N leg efflux (2.64 +/- 0.47, p less than .001) was reduced. Moreover, considering the effect of calorie load, patients receiving 30 kcal/kg.day showed a lower efflux of total AA N and of some AA considered as markers of muscle protein catabolism, such as phe, lys, met, and glu. The main difference between solutions was in the efflux of BCAA; particularly, val and leu efflux was turned into uptake in the BCAA group. No significant difference among the groups was found in N balance and MEH excretion during treatment. In brief, muscle catabolism was reduced in an amount dependent on glucose and insulin load, but it was not influenced by BCAA supply. Whole body net protein catabolism was reduced through different mechanisms, either an increased visceral N retention or a decreased muscle N loss. However, muscle N loss was never abolished even in the high calorie groups.