1. The rate of protein synthesis in the whole body was measured in one fed subject with seven 15N-labelled amino acids (intravenous and oral doses) and two 15N protein mixtures (oral doses only). The rates were determined individually from the urinary excretion of ammonia and total urea over a 12 h experimental period.
2. Except with oral glycine and alanine, the synthesis rates given by ammonia and urea were appreciably different within each study when calculated on the assumption of a single pool of metabolic nitrogen in the body. In general, intravenous administration of the tracers gave higher rates with urea and the oral route gave higher rates with ammonia.
3. The differences between intravenous and oral doses of 15N could be reduced significantly by calculating synthesis rates from either the arithmetic or harmonic average of flux rates given by ammonia and urea. The averages correspond to estimates of the total flux in a two-pool model of metabolic nitrogen when it is assumed either that both pools receive an equal amount of tracer (arithmetic) or that both have the same rate of nitrogen turnover (harmonic).
4. By so reducing the effect of physical separation of nitrogen in the body, the metabolic aspects of compartmentation of the tracer could be examined. The results show that the absolute value obtained for protein synthesis depends on the source of labelled nitrogen. The data are discussed in this empirical context.