Abstract
Ethnic residential segregation is usually investigated using a constrained-choice approach. This study explains the variation in post-war Afro-Caribbean segregation in fifteen British cities by means of historical patterns of economic opportunity. Its dependent variable is newly available census data on residential segregation. It finds that the observed variation in segregation levels cannot be explained in terms of council housing policies or the passage of civil rights legislation from the mid-1960s, but rather by the interaction of New Commonwealth immigration and local labour and housing market conditions during the critical period 1951-1966.