The authors construct a regional model of residential segregation in the South, using demographic and socioeconomic variables. Based on an analysis of the age, size, percent black, and occupational and income differentials in 32 southern cities, the findings show that: (1) age is still the strongest predictor of residential segregation; (2) socioeconomic factors need to be examined in their traditional southern context as indicators of residential segregation; (3) occupational—not income—differences have the greater effect on levels of residential segregation; and (4) changes in socioeconomic status for blacks in the South have not been accompanied by concomitant changes in residential segregation. Problems of developing and testing causal models of segregation are discussed, combined with specific suggestions for future research.