Literature regarding transit?s impact on land values reports mixed results concerning the economic benefits of accessibility to subway stations, specifically regarding commercial properties. After examining 731 commercial land values in Seoul, Korea, this study suggests a possible explanation for the mixed results: transit?s discrimination impact on land values by location in a built-up urban area. The regression coefficient for distance to station in the central business district is the highest, the subcenters are next, and other areas are lowest ? apparently a strong correlation with higher centrality and development densities of submarkets. Also, the inclusion of spatial lag and error term variables greatly improves the goodness of fit of the regression equations lowering the spatial autocorrelation in the ordinary least squares residuals as well as reduces overestimation of value premiums in association with rail transit stations, which enables a regression model to produce a more accurate and efficient estimator for transit?s impact on commercial land values.