Thalidomide as an effective treatment for multiple myeloma and leprosy has also caused birth defects in thousands of children five decades ago particularly in Europe. Thus its use in humans remains limited. The rapid and fatal approval of thalidomide at that time ultimately was a consequence of the sole use of thalidomide-insensitive species in animal toxicity tests. Here, we aimed at elucidating the molecular basis for the resistance of mice to thalidomide teratogenicity. By using hydroethidine staining we demonstrate that thalidomide induces the formation of superoxide in embryonic fibroblasts of thalidomide-sensitive species but not in those of mice. As determined by trypan blue staining, scavenging of superoxide prevents thalidomide-induced apoptosis, a marker for thalidomide teratogenicity. Mouse embryonic fibroblasts are found to have higher glutathione levels than those of sensitive species and can be sensitized for thalidomide by glutathione depletion with diethyl maleate or diamide. Accordingly, experimental increase of glutathione levels in human embryonic fibroblasts by adding N-acetyl cysteine or glutathione ethyl ester to the culture medium counteracts thalidomide-induced apoptosis. Finally, we show that thalidomide-induced molecular pathology downstream of superoxide is essentially identical in human and sensitized mouse embryonic fibroblasts. In conclusion, thalidomide-resistance is based on the capacity of the glutathione-dependent antioxidant defense. We provide a basis to pharmacologically overcome the limitations of thalidomide use at humans and describe substantial differences between human and mouse embryonic cells regarding the protection against oxidative stress.