The capacity of human plasma to bind specifically dihydrotestosterone (DHTBC) has been determined in normal monozygotic and dizygotic twins less than 20 years old. The total variation in DHTBC was analysed by stepwise multiple regression, combined with variance analysis, applied to the pair sums in DHTBC, body height and body weight, as well as to age. A significant part of the variation in DHTBC was explained. Due to the strong interwining between body weight, body height and age in growing children it was impossible to make out which one of the latter variable contributed most to the negative correlation which was observed. Previous studies in adults did, however, show that only body weight and not body height or age correlated negatively with DHTBC. By doing the same calculations for the within-pair differences in DHTBC, body height and body weight it was hoped to delineate the within-family part of the environmental variation in DHTBC. Body weight and body height did, however, not offer any significant explanation as to this part of the variation in DHTBC. The upper limit of heritability (h2) for DHTBC was estimated by comparing intraclass-correlation coefficients in monozygotic and dizygotic twins. According to the method used to calculate the latter intraclass-correlation coefficients h2 estimates for DHTBC of 0.774 and 0.960 were arrived at. However, due to the rather small number of dizygotic twins available the confidence limits of these estimates were rather large.