Traditional models of local labor market racial inequality have emphasized regional location, differences in economic and social organization, and human capital differentials as explanatory factors. Such models, however, were developed before the rise of the Sunbelt, before the transformation from a goods-producing to a service-based economy was seriously under way, and before women were a substantial portion of the United States labor force. In this article we investigate whether traditional predictors successfully explain male and female racial inequality across the 100 largest MSAs in 1980. We argue that previously developed perspectives, emphasizing singular causes, lead to misspecified models that fail to capture the diversity of factors that currently impact inequality. We show how the resource approach to economic segmentation (Hodson&Kaufman 1982; Hodson 1983) provides a conceptual framework suggesting variables that more adequately explain racial inequality in major metropolitan areas, particularly for males.