This study aims to provide empirical evidence for two traditional research questions in the field of telecommuting studies: (1) Does telecommuting promotes dispersion of urban space? (2) Does telecommuting substitute for household travel? Although these causality issues have received great deal of attention, no multivariate analysis approaches exist. Using the 2006 household travel survey data from the Seoul Metropolitan Area, this study adopts a path analysis to discover the complex relationships between telecommuting, residential/job locations, and household travel. First, the path analysis shows that rather than telecommuting serving as the determinant of location choice, job locations determine the choice to telecommute. Hence, secondary impacts of telecommuting on travel may not occur with location changes as the medium. Second, the analysis also shows that the household head's telecommuting has a positive influence on his/her non-commuting travel in both the person kilometers traveled (PKT) and vehicle kilometers traveled (VKT) models and on household members' travel in VKT models. Moreover, the VKT model suggests that the household head's non-commuting travel has a negative impact on the household members' travel. These results indicate that although telecommuting reduces commute travel, this may be offset by other travel demand within the household, owing to exhaustion of the limited travel budget. Thus, planners and policymakers must consider these impacts when evaluating the benefits and costs of telecommuting as an urban management policy. [ABSTRACT FROM AUTHOR]