The black-white gap in low birth weight in the U.S. remains large and mostly unexplained. We explore the relationship between racial residential segregation and birth weight for the period 1970-2010. A negative effect of segregation on black birth outcomes emerges only after 1980. Controls for maternal socioeconomic status and behaviors account for between 35-40 percent of the full segregation effect. Single-motherhood and mother’s education, and unobservable factors that load onto these variables, play important and increasing roles. After controlling for MSA and parent characteristics, segregation explains 21-25 percent of the black-white gap in low birth weight between 1990-2010.