This paper analyzes the link between population suburbanization and racial residential segregation in U.S. metropolitan areas. I hypothesize that rapid suburbanization between 1960 and 2000 has caused an increase in racial residential segregation in U.S. metropolitan areas. Earlier studies fail to identify the direction of causality in explaining the relationship between suburbanization and residential segregation. Using the 1947 national interstate highway plan as an instrument for suburbanization and decennial Census data from 1960 to 2000, I find that central city population decline (suburbanization) causes racial residential segregation in a metropolitan area to rise. Estimation results from both long difference and panel regressions are robust to an array of specifications. Improvement in transportation facility reduces the cost of mobility, thereby enabling the affluent whites to move to the suburbs. As a result we observe a distinctive location pattern with nonblacks living in the suburbs and blacks in the central city of a metropolitan area. Had there been no suburbanization between 1960 and 2000, racial residential segregation on average would have declined by about 4 percentage points more than what is observed in the data. JEL Classification: R11, R23, R40, H31