OBJECTIVES: Various methods are used to monitor safety of drug exposures in pregnancy, including prospective pregnancy registries, longitudinal database studies, and case-control studies. This abstract seeks to evaluate these approaches and examine impact on interpretation of results. METHODS: Study designs for the three types of pregnancy monitoring studies were examined to determine strengths and limitations in effectively monitoring drug safety in pregnancy and the types of conclusions that can meaningfully be drawn from each. RESULTS: Pregnancy registries are prospective studies with active data collection from reliable sources providing detailed, quality data on exposure and outcome. Their voluntary nature can lead to selection bias and a lengthy enrollment period may be needed to reach sufficient sample size/power. Longitudinal database studies are population based and offer timely and cost-efficient results, but often require complex linkage of multiple databases. Misclassification bias can occur because critical variables (exposure and outcome) may lack precision/detail. Case-control studies begin with the outcome of interest and retrospectively assess drug exposure offering substantial statistical power to identify teratogens among relatively rare exposures, but are subject to recall bias. Only registries are useful in monitoring safety early in the drug lifecycle. Pregnancy registries can detect major teratogenic effects, but have limited power to detect risk of specific birth defects. Database studies can detect major teratogenic effects and due to their large size/power may also detect risk of specific defects. Both approaches can assess other outcomes. Case-control studies provide insights into the strength of association between exposure and a specific defect, but cannot be used to estimate birth defect prevalence or other outcomes. CONCLUSIONS: All three approaches are viable for studying drug safety in pregnancy, but differ in the types of conclusions that may be drawn. In choosing a design, it is important to consider study objectives and outcomes of interest.