The risk assessment of drugs in pregnancy can be improved through better knowledge and interpretation of significant experimental findings. Embryo-fetal effects can be considered as the endpoint of importance, after excluding that their occurrence is related to maternal toxicity. If embryo-fetal toxic effects occur in the presence of maternal toxicity, the dam influence should be clearly assessed, and can be taken into account in the determination of a safety margin. A determination of the lowest developmental toxic dose, if any, apart from any maternal toxicity, should be made. With positive animal developmental findings, interpretation should take into account various parameters including the incidence and the severity of the embryo-fetal insult, interspecies reproducibility, the level of exposure, metabolic pathways, and the results of possible mechanistic investigations. Hierarchization of the experimental findings should take into account all these parameters together to provide support for the risk assessment in humans.