Traffic congestion, car dependence, and environment pollution have become the key factors of sustainable urban development in China. Smart phone app was used to collect oneweek commuting data for 286 residents at the trip level in Xi′an city, and environment audit was made to collect the building environment data near their neighborhood and work locations. Using multinomial Logit regression, the impact of the traditional built environment “5D” factors and other built environment factors on the six commute modes was analyzed.The results show that the following factors have significant impact on active commute mode and public transit commute mode, that is, population and building density of the workplace, distance between workplace and city center, population and road network density of the residence, number of bus lines, and whether there are colleges or stores nearby.The built environment of the workplace has greater impact on commute mode choice than that of resident locations.In addition to the traditional "5D" factors, whether there are colleges or stores nearby has significant impact on some commute modes. Besides, personal health status has a significant impact on the commute mode choice, and obese people are more inclined to commute by car. Based on all of the above findings, suggestions of planning strategies are put forward.

built environment; commute mode; smart phone app; urban planning