Retinoic acid (25 mg/kg) administered by gavage to rats at 216 hours of gestation killed almost all conceptuses by 288 hours and all by full term. The embryolethal dose50 was 12.3 mg/kg. Embryos died from damage to the allantois leading to agenesis or hypogenesis of the chorioallantoic placenta. Suppression of cell division in, disturbed fluid entry into, and impaired normal vascularization of the allantois underlay the abnormality, the predominant element depending principally on when exposure occurred. Hydremia (Br. J. Exp. Pathol., 57:525-541, '76) affected over 50% of the sacs. Relating data from the literature to resorption patterns suggested that 25 mg/kg of retinoic acid raised a lethotoxic level (approximately 2 micrograms/ml) of retinoate in the plasma for about 5 hours. This, together with asynchronous development, was used to help explain why groups of embryos responded to the teratogen for 18 hours longer than single embryos and why exposure 18 hours before the allantois on average appears, killed some young. Comparison showed that retinoic acid was 4.8 times as embryolethal as retinol. Otherwise, each behaved qualitatively similarly, suggesting that either retinol acts through retinoic acid or via a common path entered independently by each. Placental agenesis is well documented in Soviet literature. It is almost unknown in Western teratology even though it is probably the most important cause of embryonal death following teratogenic procedures around the time of placentation.