<p>In order to deal with the urban transport challenges resulted from the upcoming aging problems in China, a random forest and Bayesian network approach is proposed to explore the impacts of the built environment on the Urban Trip Characteristics of the Elderly (UTCE). On the basis of a travel survey for the elderly conducted in Beijing, random forest models are utilized to select the built environment variables and the sociodemographic variables which are significantly associated with the UTCE. Furthermore, Bayesian networks are applied to discover the causal correlations between the selected variables and the UTCE. Through changing one or more of these built environment variables separately, the corresponding individual effects and the interaction effects are investigated respectively as the scenarios analyses. The results indicate that urban transport infrastructures and urban land uses have significant impacts on the UTCE. Improving urban transport infrastructures and enriching the travel options of the elderly are able to make the trips of the elderly more convenient. Moreover, introducing more living and entertainment facilities around the residential areas of the elderly is rational and effective to guide the elderly traveling during off-peak hours and avoid them traveling across areas.</p>