To evaluate the reliability of quantitative extrapolation of radiation-induced chromosomal damage from somatic cells to germ cells, data on the effects of several biological and physical factors on the chromosomal radiosensitivity of blood lymphocytes and stem-cell spermatogonia have been collected from the literature. The results show that most of the factors considered, such as chromosomal constitution, age, genetic constitution, species, sampling time and dose fractionation, had differential effects on the induction of chromosomal aberrations in both systems. These differential effects can easily be explained in terms of the biological differences between in-vitro-stimulated peripheral blood lymphocytes and stem-cell spermatogonia. It is concluded that only direct experiments on germ cells of higher primates and man can be used for a quantitative estimation of human genetic radiation risks arising from structural chromosomal aberrations.