Several countries have experienced rises in cryptorchidisms, hypospadias and testicular germ cell cancer. The reasons for these trends are largely unknown, but Skakkebaek has proposed that these disorders form a testicular dysgenesis syndrome and can be traced to androgen insufficiency in foetal life. This suggests that antiandrogenic chemicals might contribute to risks, but few chemicals have been linked to these diseases in epidemiological studies. In animal studies with p,p'-dichlorodiphenyldichloroethylene, effects typical of disruptions of male sexual differentiation became apparent when the foetal levels of this androgen receptor (AR) antagonist approached values associated with responses in in vitro assays. This prompted us to analyse whether the 22 chemicals with AR antagonistic properties would produce mixture effects in an in vitro AR antagonism assay when combined at concentrations found in human serum. Other antiandrogenic modalities could not be considered. Two scenarios were investigated, one representative of average serum levels reported in European countries, the other in line with levels towards the high exposures. In both situations, the in vitro potency of the 22 selected AR antagonists was too low to produce combined AR antagonistic effects at the concentrations found in human serum, although the high exposure scenario came quite close to measurable effects. Nevertheless, our analysis exposes an explanation gap which can only be bridged by conjuring up as yet undiscovered high potency AR antagonists or, alternatively, high exposures to unknown agents of average potency.