Assessment of the potential developmental toxicity of arsenic in humans must be based entirely on the extensive animal literature; no appropriate human data are available. Hazard identification of developmental toxicity of arsenic in animal studies is complicated by the co-occurrence of maternal and developmental toxicity when the pregnant dam is exposed to the toxicant. Current regulatory guidance requires that, when maternal and developmental toxicity occur at the same or similar doses, detailed consideration needs to be given to whether developmental toxicity is secondary to maternal toxicity or whether it represents a distinct hazard. In this review, these principles were applied to the relatively large database of animal studies available for hazard identification of inorganic arsenic as a developmental toxicant. It is concluded that maternal and developmental toxicity occur in the same dose range for this potent cytotoxicant, although differential no observed adverse effect levels can be identified depending on the endpoints used. Various evidence from the basic science literature indicates that developmental toxicity is not secondary to maternal toxicity. Current regulatory guidance falls short of defining effective approaches to resolving the difficulties posed by co-occurrence of maternal and developmental toxicity.