Welfare-to-work transportation programs are permissed on a conceptualization of the spatial mismatch hypothesis that focuses on the physical seperation between the central city locations of welfare participants, rapidly expanding job opportunities in the suburbs, and the long commutes needed to connect them. Using data from three diverse California counties, this study examines welfare recepients' spatial access to employment. The study finds that the traditional notion of spatial mismatch is less relevant to welfare recipients, many of whom live in countries in which the urban structure deos not fit the simple model of poor, central-city neighborhoods and distant, job-rich suburbs. Many welfare recepient live in job-rich areas; others live in neighborhoods that are spatially isolated from employment. To be effective, thereofre, transportation policies must be tailored to the diverse characterisitcs of the neighborhoods in which welfare recepients live.