Abstract: Restricting privates car use and promoting pure electric vehicles use are two important ways to realize green travel in China cities． In order to analyze the impact of travelers' environmental awareness on green travel， travelers were categorized with the latent class analysis based on travelers' attitude towards environment and trave- ling behaviors． Then travelers' latent class was introduced into the traditional bi-variate probit model as explanatory independent variables． And the impact of travelers' latent class on the approval of private car restriction policies and the purchase intention of pure electric vehicles were analyzed． The results show that according to their attitude towards environment and traveling behaviors，travelers can be categorized into three latent classes: the environ- mentalist，the non-environmentalist，the pseudo-environmentalist． There is a higher degree of fitting when incor- porating travelers' latent class into the bivariate probit model． And travelers' latent classes have great impacts on their green traveling choice．

Key words: traveler, green travel behavior, heterogeneity, latent class analysis, bivariate probit model, green transport