Results of previous studies of effects of drinking during pregnancy on standardized tests of infant development have varied, depending on level of exposure. This study of 382 Black, inner city infants confirmed effects of moderate-to-heavy prenatal alcohol exposure on Bayley Scale performance originally observed in a White, middle class cohort. Effects of second and third trimester drinking were as strong or stronger than those of drinking at the time of conception. New methods for reducing the Bayley Scale data suggested specific deficits relating to emergence of the ability to imitate modeled behavior and the development of standing and walking. The incidence of very poor performance (bottom 10th percentile) on the Bayley Mental Development Index more than doubled in children whose mothers averaged at least 0.5 oz absolute alcohol per day during pregnancy, indicating that this exposure has a clinically meaningful impact at markedly lower levels than those associated with full fetal alcohol syndrome. Clinically meaningful effects on the Psychomotor Development Index, by contrast, were seen only in the infants of heavy drinking mothers with overt symptomatology of alcohol abuse. Adverse effects on Bayley Scale performance were more prevalent in the infants of women more than 30 years of age.