Valproic acid use during the first trimester of pregnancy has been reported among an unusu ally high proportion of mothers of infants with spina bifida. During 1976 and from 1978 through September 1982, the birth defects surveillance system at the Institut Europeen des Genomutations in Lyon, France, ascertained 146 cases of spina bifida aperta. Among these cases, nine (6.2%) of the mothers had epilepsy and had taken valproic acid during the first trimester at dosages between 400 mg and 2,000 mg per day. Five of the nine patients with spina bifida were exposed to valproic acid alone, and four were exposed to additional anticonvulsants. Twenty-one (0.32%) of the mothers of the 6,61 6 infants in the surveillance system with other malformations had taken the drug (Table 1). These data show a highly statis tically significant odds ratio of 20.6.* To isolate the effect of valproic acid from the possible ef fects of seizure disorders and other drug therapy, the analysis was then confined to the 71 epileptic mothers. Nine (90%) of the 10 such mothers of spina bifida infants had taken valproic acid, compared with 21 (34.4%) of the 61 mothers of infants with other defects (Table 2). The odds ratio of 1 7.1 is statistically significant. Reported by E Robert, MD, Institut Europeen des Genomutations, Lyon, France; Epidemiology Development Br, Div o f Drug Experience, Food and Drug Administration; Birth Defects Br, Chronic Diseases Div, Center for Environmental Health, CDC. Editorial Note: A woman who requires treatment for epilepsy during pregnancy is at increased risk of having a baby with a birth defect. The American Academy of Pediatrics Committee on Drugs offers the following recommendation on alerting women to the risk: "When a woman who has epilepsy and requires medication asks about pregnancy, she should be advised that she has a 90% chance of having a normal child, but that the risk of congenital malformations and mental retardation is two to three times greater than average because of her disease or its treatment" ( 1). The new data from Lyon do not change this general advice.