Many US cities invest in large public transit projects in order to reduce private vehicle dependence and to reverse the downward trend in public transit use. Using a unique panel data set for five major cities that upgraded their rail transit systems in the 1980s, we estimate new rail transit's impact on usage and housing values, using distance as a proxy for transit access. New rail transit has a small impact on usage and housing values. This impact is enough to represent tangible benefits of new transit to nearby residents. New transit's benefits are not uniformly distributed. We document which demographic groups are over represented in transit growth areas and the changes in transit usage by different demographic groups. (C) 2000 Elsevier Science S.A. All rights reserved.