Abstract: The subway Automatic Fare Collection systems collect tremendous amount of smart card data, which provides comprehensive spatial-temporal information about subway passengers. The analysis of passengers travel patterns benefits urban rail transit operation companies in predicting subway passenger flow and formulating operational strategies. A data-mining procedure to identify travel patterns of subway passengers was introduced: after pretreatment of smart card data, passengers travel chains were generated based on the spatial-temporal information of it; clustering variables that reflect spatial-temporal characteristics of passengers were analyzed; K-means clustering algorithm was adopted to cluster the passengers; the potential passengers travel patterns were then analyzed. Taking the Shenzhen subway smart card data as example, verification experiment was conducted on the proposed subway passengers traveling patterns analysis methodology.