Endocrine disruptors (ED) have been incriminated in the current increase of male reproductive alterations. Bisphenol A (BPA) is a widely used weak estrogenic environmental ED and it is debated whether BPA concentrations within the average internal exposure are toxic. In the present study we investigated the effects of 10(-12) to 10(-5) M BPA concentrations on fetal Leydig cell function, as fetal life is a critical period of sensitivity to ED effects on male reproductive function. To this aim, fetal testes from human at 6.5-10.5 gestational weeks (GW) or from rat and mouse at a comparable critical period of development (14.5 days post-coitum (dpc) for rat and 12.5 dpc for mouse) were explanted and cultured using our validated organotypic culture system in the presence or absence of BPA for 1-3 days. BPA concentrations as low as 10(-8) M reduced testosterone secretion by human testes from day 1 of culture onwards, but not by mouse and rat testes where concentrations equal to 10(-5) M BPA were required. Similarly, 10(-8) M BPA reduced INSL3 mRNA levels only in human cultured testes. On the contrary, 10(-5) and 10(-6) M diethylstilbestrol (DES), a classical estrogenic compound, affected testosterone secretion only in rat and mouse testis cultures, but not in human testis cultures. Lastly, contrarily to the DES effect, the negative effect of BPA on testosterone produced by the mouse fetal testis was maintained after invalidation of estrogen receptor α (ERα). In conclusion, these results evidenced i) a deleterious effect of BPA on fetal Leydig cells function in human for concentrations from 10(-8) M upwards, ii) species-specific differences raising concerns about extrapolation of data from rodent studies to human risk assessment, iii) a specific signaling pathway for BPA which differs from the DES one and which does not involve ERα.