This article reviews the difficulties in studying adverse effects of drugs during pregnancy on the fetus and newborn. A study design should strive for prospective recording of drug intake during pregnancy, comparison to appropriate control groups adjusted for inherited traits, and single drug exposures for evaluation of specific syndromic causation, such as the Fetal Hydantoin Syndrome. Animal models are best used in mechanistic study of adverse drug effects on the fetus rather than for screening for adverse effects. Careful conclusions about causation of drug-induced adverse effects are needed to avoid false associations while providing appropriate safeguards to the fetus and newborn.