In vivo measurement of muscle protein synthesis and its hormonal regulation is limited by the difficulty of measuring aminoacyl-tRNA specific activity (SA). We assessed the kinetics of heart and skeletal muscle phenylalanyl-tRNA labeling during continuous infusion of L-[ring-2,6-3H]phenylalanine (Phe) to fasted anesthetized rats. We measured Phe SA in arterial and femoral venous plasma, the tissue acid-soluble pool and muscle protein hydrolysates after 5 min (n = 7), 30 min (n = 6), and 90 min (n = 7). We also assessed insulin's effect on labeling of the tRNA pool and muscle protein synthesis during a hyperinsulinemic clamp (2 mU.kg-1.min-1; n = 7). Labeling of tRNA in heart reached 59 +/- 5, 67 +/- 3, and 83 +/- 3% of arterial SA at 5, 30, and 90 min of saline infusion, respectively, but only 10 +/- 5, 34 +/- 2, and 48 +/- 2% in skeletal muscle at those times (P < 0.01 vs. heart). The tRNA SA was intermediate between SA in the acid-soluble pool and arterial plasma. Femoral venous SA was 32 +/- 2% lower (P < 0.001) than arterial SA. Skeletal muscle tRNA SA was also 29 +/- 3% lower (P < 0.001) than femoral venous SA. Insulin did not alter tRNA labeling and neither heart (9.8 +/- 1.1%/day for saline vs. 8.4 +/- 1.0%/day for insulin) nor skeletal muscle (6.7 +/- 1.5%/day vs. 4.2 +/- 0.4%/day) protein synthesis. Thus labeling of phenylalanyl-tRNA occurs more rapidly in heart than in skeletal muscle and is unaffected by insulin.(ABSTRACT TRUNCATED AT 250 WORDS)