To investigate the relationship between maternal cocaine dependency and child maltreatment in a cohort of young children. Historical cohort study at an urban, tertiary care medical center. 47 infants, born between January and September 1989, whose mothers were regular users of cocaine during pregnancy, based on history and the results of newborn's urine toxicology screens. These cocaine-exposed infants were matched to a comparison group of 47 infants whose mothers did not use cocaine during pregnancy. Matching was on the basis of birth date, race, method of payment for the hospitalization, and marital status of the mother. Occurrence of maltreatment (physical abuse, sexual abuse, or neglect), and placement either in foster care or with a substitute caretaker. By 24 months of life, maltreatment had occurred in 23% of the cocaine group vs 4% of the comparison group (risk ratio, 5.5; 95% confidence interval, 1.3 to 23.5). Physical abuse had occurred in 11% of the cocaine group vs 2% of the comparison group, while neglect had occurred in 11% vs 0% (P < .05). Changes in placement had occurred in 20% of the cocaine group vs 2% of the comparison group (risk ratio, 10.0; 95% confidence interval, 1.3 to 75.1). Of the 10 placements, only three were directly linked to an episode of maltreatment. Children identified during the neonatal period as regularly "exposed" to cocaine in utero are at a substantially increased risk both of maltreatment and of changes in the primary caretaker during the first 24 months of life.