The goal of this study was to discover whether using different tracers affects the measured rate of muscle protein synthesis in human muscle. We therefore measured the mixed muscle protein fractional synthesis rate (FSR) in the quadriceps of older adults during basal, postabsorptive conditions and mixed meal feeding (70 mg protein x kg fat-free mass(-1) x h(-1) x 2.5 h) by simultaneous intravenous infusions of [5,5,5-(2)H(3)]leucine and either [ring-(13)C(6)]phenylalanine or [ring-(2)H(5)]phenylalanine and analysis of muscle tissue samples by gas chromatography-mass spectrometry. Both the basal FSR and the FSR during feeding were approximately 20% greater (P < 0.001) when calculated from the leucine labeling in muscle tissue fluid and proteins (fasted: 0.063 +/- 0.005%/h; fed: 0.080 +/- 0.007%/h) than when calculated from the phenylalanine enrichment data (0.051 +/- 0.004 and 0.066 +/- 0.005%/h, respectively). The feeding-induced increase in the FSR ( approximately 20%; P = 0.011) was not different with leucine and phenylalanine tracers (P = 0.69). Furthermore, the difference between the leucine- and phenylalanine-derived FSRs was independent of the phenylalanine isotopomer used (P = 0.92). We conclude that when using stable isotope-labeled tracers and the classic precursor product model to measure the rate of muscle protein synthesis, absolute rates of muscle protein FSR differ significantly depending on the tracer amino acid used; however, the anabolic response to feeding is independent of the tracer used. Thus different precursor amino acid tracers cannot be used interchangeably for the evaluation of muscle protein synthesis, and data from studies using different tracer amino acids can be compared qualitatively but not quantitatively.