Study of 284 births in Rochester, Minn, to 138 women who had epilepsy suggests (as have other studies) that there is an increased incidence of malformations among the offspring of women taking anticonvulsants. Among the 141 children born to mothers taking one or more of the anticonvulsants, ten had serious malformations, including five with congenital heart lesions, two with cleft palate, and one with both heart and palate anomalies. Among 56 offspring to mothers with epilepsy who did not take anticonvulsants in the first trimester but were comparable with the 141 in respect to age, parity, and type of seizure, there was one malformation (agenesis of the testis). There were no malformations among 26 offspring of mothers whose epilepsy had been in remission or among 61 offspring of mothers whose pregnancies predated the epilepsy.(Aren Neurol 31:364-373, 1974) Evi dence has been accumulating during the last few years that some anticonvulsant medications may have a teratogenic effect. These are among the more commonly used of all drugs, for epilepsy is found in_ approximately 0.5% of all populations Purveyed and most patients with ep¬ ilepsy use one or more anticonvulsant medications. These drugs include the following: phénobarbital, introduced in 1912; diphenylhydantoin (Dilantin), introduced in 1938 and since then widely used, often in combination with phénobarbital; and, in approxi¬ mate order of use in the United States, mephobarbital, mephenytoin (Mesantoin), and primidone (Myso¬ line). Other anticonvulsants are used less often. In this communication, we examine the literature on anticonvulsants and teratogenicity and present the results of a survey of congenital malforma¬ tions among the children of women in whom epilepsy was diagnosed at the Mayo Clinic and who were delivered at a hospital in Rochester, Minn. Problems in the Evaluation of Epidemiologie StudiesIn the epidemiologie study of the possible teratogenic effects of anti¬ convulsants, several important prob¬ lems require consideration at the out¬ set. A major difficulty isj;hej^ejative]y _ low preyalenee-of epilepsy^about-fliie in_2ÜÍLpersons) and of congenital mal¬ formations (about 20 to 30_/^000. births). If fertility among individuals with epilepsy is the same as that in the general population,_one woukLex-_ pect five in each 1,000 births to be of Q mothers with epilepsy. Since 1940, at the Mayo Clinic 3.9/1,000 births were to women with active epilepsy. If, further, the incidence of congenital malformations is t/rh same as 1 . the.general pspulation,_we.w_q.u -rjte£tanly-Qna.in KUMLhirths to .he a .malformed child born ±o a woman with epilepsy.