The validity of using different 15N-tracer substances to measure whole body protein parameters, i.e., protein synthesis, protein breakdown, net protein gain, protein turnover, metabolic pool, and reutilization, was assessed by comparing the results obtained with: [15N]glycine, a mixture of 10 15N-labeled amino acids, and a 15N-labeled chicken egg protein in two infants, 9 and 12 weeks old, who were fed human milk. The tracer substances were fed orally as a single dose corresponding to a 15N-excess quantity of 0.2 mmol X kg-1 body weight. 15N Excretion in the urine was measured cumulatively by emission spectrometry, and the data on the protein metabolism were calculated by means of a three-pool model. All three tests yielded consistent net protein gains. The protein synthesis, protein breakdown, protein turnover, and nitrogen reutilization values produced by the [15N]glycine tracer study were higher than those produced by application of the 15N-amino acid mixture and the 15N-labeled egg protein. However, in our opinion, this discrepancy does not justify the replacement of [15N]glycine by expensive 15N-amino acid mixtures as tracer substances.