BACKGROUND: Hypospadias is one of the most frequently occurring genital anomalies described in infants prenatally exposed to valproic acid (VA). However, to our knowledge, only one publication has studied a potential causal relationship between VA and hypospadias, only estimating the unadjusted global risk. Here we present the results of a multivariate case-control study aimed at analysing and quantifying the specific risk of hypospadias in newborn infants exposed to VA during the first trimester of pregnancy. METHODS: The data analysed here were derived from the Spanish Collaborative Study of Congenital Malformations (ECEMC), an ongoing, hospital-based, case-control study and surveillance system in which collaborating paediatricians identify case and control infants. The paediatricians collect the same data for both case and control infants, blinded to information on any prenatal exposure. The information includes 312 items related to many prenatal exposures, including drug exposure, reproductive and family history, and other characteristics. The sample analysed included 2,393 infants with hypospadias and 12,465 male controls. RESULTS: The results showed that the unadjusted risk of hypospadias in infants prenatally exposed to VA was 5.23 (95% CI 2.31, 11.86; p < 0.00001). Once adjusted for 13 potential confounding factors using conditional logistic regression analyses, the value of the risk was of a similar magnitude (odds ratio = 5.71; 95% CI 1.78, 18.36; p = 0.003). In addition, the frequency of hypospadias in the study population was approximately 1.8/1000 births. This allowed us to calculate the specific risk for an infant with hypospadias to be born to an exposed mother, which was 1 child in 97 births to mothers using VA during the first trimester of pregnancy. We consider this information much more useful for risk assessment than the risk value itself. CONCLUSIONS: An alteration of placental gonadotrophic stimulation caused by changes in gonadotropin-releasing hormone release produced by the effects of VA on GABA is a possible pathogenic mechanism. Our results support the relationship between prenatal exposure to VA and hypospadias.