This study aimed to evaluate how built environments, especially land use, affect time-of-day transit demands, using data collected from neighborhoods around subway stations in Seoul, South Korea. The stations were first clustered by their boarding and alighting ridership sequences across times. The outcomes showed that the geographical distribution of clustered stations coincides with land use divisions. For each cluster, we estimated latent growth curve models to examine how land use determines time-of-day ridership patterns. The modeling outcomes showed that the same land use pattern and intensity differently affect ridership volumes across times and locations. The findings in this study indicated that uniform land use in areas is a dominant factor for asymmetric ridership patterns-passengers are inevitably concentrated in peak hours if a certain land use is predominant-and these asymmetric ridership patterns can be mitigated when neighborhood commercial land use is included in the areas.