This paper explores the role of maternal drug use and the timing of prenatal care. The study data were collected from women delivering live births at eight participating hospitals in the Washington, D.C., Metropolitan Area Drug Study. An estimated 16.9 percent of the women in this sample initiated prenatal care in their third trimester or received no prenatal care. After adjusting for age, race/ethnicity, education, parity, and attitude toward pregnancy, cocaine use was strongly associated with the timing of prenatal care. Using multivariable ordinal logistic regression, the data suggest significant barriers to prenatal care for substance abusers, especially cocaine users. Increasing access to prenatal care continues to be an important public health policy objective, particularly in urban areas where substance abuse is prevalent. Health services research must test strategies that address the timing of prenatal care among drug-dependent, urban women.