Generalized estimating equations were used to examine the relation of low birthweight, an indicator of adverse perinatal events, and social disadvantage to IQ changes during the period of school attendance. Data are from a longitudinal study of low birthweight and normal birthweight children in two disparate communities, an inner-city and near-by suburbs in southeast Michigan (n=773). Wechsler intelligence tests were administered at ages 6, 11 and 17. Low birthweight-related deficits (vs. normal birthweight) detected at the start of schooling were about 5 IQ points and these remained constant up to age 17. Initial IQ deficits associated with urban environment (vs. suburban) increased significantly from age 6 to 11, but no further by age 17. These trends were independent of one another: The low birthweight deficit was constant across social environments; the social disadvantage deficit was uniform across birthweight groups. The finding that the urban--suburban gap did not continue to widen after age 11 probably resulted from an atypical IQ decline of suburban children. The causes of this unexpected finding are unclear.