The effect of breast-feeding on intellectual development remains controversial. We explored this relationship in a high socioeconomic population in which breast-feeding was supplemented with soy containing formulas at some time during the first year of life. As part of the 1988 National Institute of Child Health and Human Development school-based survey of two metropolitan Washington, D.C. counties to identify children in the 1978 to 1979 birth cohort who had been exposed to the chloride deficient formulas Neo-Mull-Soy and Cho-Free during infancy, information on breast-feeding was also obtained on children exposed to the chloride-deficient formulas and a group of control children exposed to other soy formulas. Because no differences in intellectual development were observed between the two groups, they were combined and the effect of breast-feeding on intellectual development at 9 and 10 years was assessed. There were 176 infants that received no breast-feeding and 342 who were breast-fed. The median duration of breast-feeding was 124 days (interquartile range, 42-248 days). There were no differences in birth weight, gender or race between the infants who were breast-fed and those who were not. The mean Weschler Intelligence Scale-Revised Full Scale IQ was 122 among those breast-fed compared to 118 among those that were not (P = 0.0008). However, following adjustment by linear regression for maternal education, paternal education and annual income the adjusted mean full scale IQ was 111 among the breast-fed and 110 among the non-breast-fed (P = 0.23). Further analyses limited to those exclusively breast-fed for the first 60 days failed to demonstrate any significant relationship between breast-feeding and IQ.