Abstract: An important methodological issue in accessibility research is how the results are affected by geographical scale. Understanding the scale effect and the associated modifiable areal unit problem (MAUP) is also important for the analysis of land use–travel interaction. Using a distinct type of accessibility measures, namely space–time measures, this study examines whether the relationships among accessibility, land use, and personal and household characteristics vary systematically with geographical scale. Space–time accessibility measures were implemented using an activity–travel diary data set collected in the Portland (Oregon, USA) metropolitan area. Through multilevel models, the study shows that these relationships are scale invariant.