This paper analyzes the effects of commuting distance on quit and move propensities. In metropolitan areas with conventional wage and housing price gradients, most workers ordinarily move in order to lengthen commutes and quit in order to shorten them. However, quits and moves by workers whose residential choices are constrained by segregation should be relatively insensitive to commutes. Descriptive statistics and simultaneous probit estimates of move and quit propensities for white and black employees of a single firm confirm these predictions. They demonstrate that long commutes encourage white quits and discourage white moves. Commute increases of one standard deviation would increase white quit propensities and reduce white move propensities by approximately 10%. In contrast, commutes by black employees have no significant effects on their quit and move propensities.