Male infertility is a major public health issue predominantly caused by defects in germ cell development. In the past, studies on the genetic regulation of spermatogenesis as well as on negative environmental impacts have been hampered by the fact that human germ cell development is intractable to direct analysis in vivo. Compared with model organisms including mice, there are fundamental differences in the molecular processes of human germ cell development. Therefore, an in vitro model mimicking human sperm formation would be an extremely valuable research tool. In the recent past, both human embryonic stem (ES) cells and induced pluripotent stem (iPS) cells have been reported to harbour the potential to differentiate into primordial germ cells and gametes. We here discuss the possibility to use human amniotic fluid stem (AFS) cells as a biological model. Since their discovery in 2003, AFS cells have been characterized to differentiate into cells of all three germ layers, to be genomically stable, to have a high proliferative potential and to be non-tumourigenic. In addition, AFS cells are not subject of ethical concerns. In contrast to iPS cells, AFSs cells do not need ectopic induction of pluripotency, which is often associated with only imperfectly cleared epigenetic memory of the source cells. Since AFS cells can be derived from amniocentesis with disease-causing mutations and can be transfected with high efficiency, they could be used in probing gene functions for spermatogenesis and in screening for male reproductive toxicity.