In this article, we argue that segregation interacts with a variety of stru
ctural transformations in society to determine the spatial concentration of
poverty. Based on this argument, we then specify a statistical model overc
oming methodological problems that have hampered earlier work. Estimates ba
sed on US data confirm that racial/ethnic segregation interacts with struct
ural shifts in society to concentrate poverty. By 1990, a powerful interact
ion between residential segregation and income inequality had emerged to sp
atially isolate the poor, an interaction the effects of which were buttress
ed by weaker interactions between segregation, rising class segregation, an
d stagnating mean incomes. Our analysis reveals how underlying shifts in so
cio-economic structure can have very different effects on the concentration
of poverty experienced by different groups, depending on the degree of rac
ial/ethnic segregation they experience.