We test the hypothesis of transit-induced gentrification for the Regional Transportation District light rail system in Denver, CO. We use a quasi-experimental spatial econometric approach, the spatial difference-in-differences model, to measure the causal relationship between urban rail investments and gentrification, which allows us to capture the average direct and indirect (spatial spillover) effects of urban rail on several socioeconomic measures of gentrification. We further account for unobserved heterogeneity and spatial dependence via the use of a panel data estimator with spatial error components. Our analysis shows that the installation of a light rail station significantly increases household income and housing values in neighborhoods up to one mile from the station.