Prior research has established associations between pregnancy outcomes and specific neighborhood characteristics, including economic disadvantage, violent crime, and racial/ethnic segregation. Recently, associations have also been found between various health outcomes and group density, the degree to which an individual is a racial or ethnic majority in his or her local community. The objective of this study was to determine the extent to which census tract economic disadvantage, violent crime rate, and group density are associated with pregnancy outcomes among White, Black, and Hispanic infants in a large metropolitan setting. This cross-sectional study utilized 1990 census data, 1991 crime data, and 1991 birth certificate information for singleton live births in Chicago, Illinois. Results show substantial racial segregation in Chicago, with 35% of census tracts having more than 90% Black residents and 45% of census tracts having fewer than 10% Black residents. After stratifying by maternal race/ethnicity, we used multilevel analyses to model pregnancy outcomes as a function of individual and census tract characteristics. Among all racial/ethnic groups, violent crime rate accounted for most of the negative association between tract economic disadvantage and birth weight. Group density was also associated with birth weight but this association was stronger among Whites and Hispanics than among Blacks. Further analysis revealed that group density was more strongly associated with preterm birth while violent crime rate was more strongly associated with small for gestational age. These results suggest that group density and violent crime may impact birth weight via different mechanisms.