The effect of the route of nutrient administration on the relative rates of leucine and phenylalanine kinetics was examined in 30 low birth weight (LBW) infants using L-[1-(13)C]leucine, L-[2H5]phenylalanine, and L-[2H2]tyrosine tracers. The infants received special premature formula (PF, n = 10, 117 +/- 8 kcal.kg-1.d-1 and 3.2 +/- 0.2 g protein.kg-1.d-1) or fortified human milk (HM, n = 10, 106 +/- 6 kcal.kg-1.d-1 and 3.0 +/- 0.2 g protein.kg-1.d-1), or parenteral nutrition (PN, n = 10, 80 +/- 25 kcal.kg-1.d-1 and 1.8 +/- 0.3 g protein.kg-1.d-1). The rate of appearance (Ra) of leucine (RaLeu), was significantly higher in group PF as compared with groups HM and PN (434 +/- 51 versus 377 +/- 33 and 359 +/- 50 mumol.kg-1.h-1, p < 0.05). The Ra of phenylalanine (RaPhe) was lower in group HM as compared with group PF (94 +/- 18 versus 115 +/- 16, p < 0.05), RaPhe in group PN (108 +/- 24 mumol.kg-1.h-1) was in between group PF and HM. The relative rate of RaPhe and RaLeu expressed as RaPhe/ RaLeu ratio was lower in all groups than that expected from reported whole body protein composition and from that reported in adults. The ratio of phenylalanine hydroxylation to leucine decarboxylation was 0.202 in group PF, 0.212 in group HM, and 0.161 in group PN, suggesting a higher rate of decarboxylation of leucine relative to hydroxylation of phenylalanine. We conclude that: 1) the higher RaLeu compared with the RaPhe may be the result of either a higher turnover of a body protein enriched in leucine or the consequence of higher leucine intake in infant nutrition and 2) whole body protein kinetics calculated from a single amino acid tracer do not adequately represent whole body protein metabolism.