Abstract: To balance the competitive relationship between the taxi and car-on-demand, a collaborative pricing mechanism of the two was developed. The sharing proportion of the two modes was determined by establishing the passenger competition utility model based on the Stackelberg model and dynamic game theory. And with a bi-level programming game model, parameters such as operating profit and user utility were used to establish the optimized pricing model of taxi and car-on-demand. Then the optimal solution of the multi-level programming model was obtained by using the sensitivity analysis method. Finally, taking Tianjin as an example, passenger's SP was investigated combined with the development status of taxi and car-on-demand. The acceptable pricing range and sharing proportion of the two modes were obtained by putting the actual data into the model when the passenger ratio was relatively stable. The research provides reference for avoiding the vicious competition between the taxi and the car-on-demand, and for promoting the coordinated development of the two, so that the two complement each other to provide users with more diversified way of travel.