Debate over how land use and self-selection affect travel behavior continues. Prior research contributes limited empirical evidence to this debate, and characterizing self-selection remains problematic. This empirical research explores the impacts of self-selection and proximity to transit at both residence and workplace. The research hypothesis is self-selection and proximity to transit increase the probability of workers commuting by rapid rail transit. To conduct this research, a station-exit passenger survey was conducted along the Taipei Rapid Transit System. Analysis methods include binomial logit modeling and sensitivity analysis. Research results support the idea that transit proximity to both work and residence increase the probability of transit commuting, but the hypothesis about the impact of self-selection is only partly supported. Policy implications suggest that, on one hand, increasing density around transit stations could realize unfulfilled self-selection; on the other hand, improved quality-of-life characteristics in neighborhoods around station areas may induce residents and companies to relocate to the neighborhood, thereby increasing residents’ and workers’ probabilities of commuting by transit.