The relationship between maternal aspirin use during the first 20 weeks of pregnancy and the child's IQ at 4 years of age was investigated in 19,226 pregnancies occurring from 1959 to 1966 in the Collaborative Perinatal Project. The mean IQ of children exposed to aspirin was 98.3, which was 2.1 points higher (95% confidence interval = 1.7, 2.6; P less than 0.0001) than that of unexposed children. Adjustment for multiple social, demographic, and other confounders reduced this difference to less than one point in favor of the aspirin exposed group, although statistical significance remained. Total days of exposure was used as an index of dose, and no dose-response relationship between aspirin use and IQ was found. The effect of prenatal aspirin exposure did not vary by infant sex. It is concluded that an adverse effect of aspirin exposure on IQ is unlikely.