This paper attempts to determine whether residential location affects unemployment duration. Our analysis is based on a spatial job search framework that shows the importance of dissociating the role of travel time from physical distance in unemployment duration. The contribution of our study also stems from the development of skill-specific accessibility measures that take into account the spatial distribution of labor supply and demand. Our results show that physical distance and competition among searchers must be controlled for in order to understand the significant role of job access (measured in terms of travel time) in unemployment duration. Second, improvements in access raise the probability that persons living in urban fringes and rural areas will become employed. Third, for workers living in large urban centers, the relationship between location and unemployment duration is insignificant.