Estimates of protein-synthesis rates using radioisotopes require accurate measurement of the specific radioactivity of the label in protein and in the precursor pool over time. Although the extracellular and intracellular pools of amino acids are easiest to sample, the tRNA pool is the direct precursor and is the appropriate pool for sampling. To test if the intracellular or extracellular pools reflect the tRNA specific radioactivity, a chicken macrophage cell line was incubated in medium containing either 0.23 mM-leucine and 14.5 microCi of [3H]leucine (tracer dose) or 2.3 microM-leucine plus 145.0 microCi of [3H]leucine (flooding dose). At both leucine levels, the tRNA specific radioactivity reached a plateau quickly, but did not equilibrate with either the extracellular or intracellular specific radioactivity within 30 min, and remained closer to that of protein. In a second experiment, proteins in chicken macrophages were labelled with [3H]leucine for 2 days. Labelling medium was removed, and the cells were washed free of residual free [3H]leucine and incubated with medium containing either 0.23 mM- or 2.3 mM-leucine (unlabelled). The specific radioactivity of leucyl-tRNA leucine reached a plateau within 2 min and remained considerably closer to that in the protein than that in intracellular or extracellular pools for at least 60 min. These results suggest that amino acids from protein degradation are a primary source for charging tRNA. When protein-synthesis rates are estimated by label incorporation, use of extracellular or intracellular specific-radioactivity values result in a marked underestimation.