The terminology to be used in reproductive (or in prenatal) toxicology has to be in accord with other fields and principles of toxicology; the reasons are briefly discussed. In addition it is essential to assess prenatal toxicity in comparison to adult (maternal) toxicity. Since pharmacokinetics in laboratory animals (e.g. rodents) usually differ considerably from that in man, this fact has to be considered when planning and evaluating studies on prenatal toxicity. Up till now this aspect has seldom been taken into account. A special problem in prenatal toxicity is the inter- and intralitter variability of the toxic manifestation (especially in polytocal animals). This problem has to be recognized by the investigators and means of dealing with it have to be developed. Like all other toxic effects, embryo-/fetotoxic manifestations occur dose dependently. Little information is available in the literature on clean dose-response-curves for teratogenic effects. Some data from our laboratory are presented. Risk assessment of teratogenic effects up till now represents a major problem. While qualitative risk assessment for man on the basis of animal data is possible, quantitative extrapolation from such data to the situation possibly existing in man is still difficult, because basic principles and strategies are largely lacking (e.g. may a "threshold" be assumed or not?). The results of some activities towards this goal are presented from our laboratory.