Multiple surveillance pregnancy registries have been established in order to better understand the effects of antiepileptic drugs (AEDs) on pregnancy. These registries are either hospital based, population based, or pharmaceutical based and are primarily focused on the potential teratogenicity of AEDs. The main outcome variable for most of these registries is the risk of major congenital malformations. Registries also gather data on other aspects of pregnancy, including seizure control. The methodology of the registries varies. They have different populations, ascertainment strategies, follow-up, and reporting criteria. These differences limit the ability to allow direct comparisons. Overall, the registry data suggest that the overwhelming majority of women with epilepsy treated with AEDs will have normal, healthy babies. A second consistent finding is that valproate, particularly at higher doses, is associated with a higher risk of major congenital malformations than other AEDs.