Experiments have been performed to validate the technique of determining total body nitrogen by measuring the induced activity of 13N produced by the reaction of 14 MeV neutrons with nitrogen. Firstly, the accuracy of determining the counts due to nitrogen was studied by measuring the activities of the positron-emitting nuclides of C, Cl, K, P, and O, as well as N, produced by the irradiation of anthropomorphic phantoms with a surface neutron dose equivalent of 0.5 mSv (50 mrem). Secondly, the effect of variation of spatial sensitivity was studied by measuring the activation and detection efficiencies of small samples at many positions within an anthropomorphic phantom, and also by comparing the known amounts of nitrogen within a new type of phantom in which the elements are fixed, with the amounts measured using a calibration based on experiments with a Bush phantom in which the elements are freely diffusible within each phantom part. There was a maximum difference of only 4% between the measured and known amounts of nitrogen. It is therefore concluded that this technique provides a simple and accurate way of measuring total body nitrogen.