The study is undertaken to explore the parking behavior of motorcycle users by using the Stated Preference Method. A two- stage approach is used to follow the principles of experimental design of Stated Preference Method. We try to explore the attitude of motorcycle users and their parking demand characteristics, and find the range of related attribute values in the first stage in order to design the levels of specific attributes. Based on the information derived from the first stage, the normal distribution method was used to design the stated preference questionnaires. A face-to-face interview was conducted to collect enough data to establish the parking behavior model for motorcycle users. Some long term, medium term, and short term parking control strategies were also derived and tested in this study.The empirical results show the tests of significance and goodness of fit for multinomial logit model are pretty well. Adding the terms of sex and income will significantly increase the explanatory ability. According to the results of policy simulation, we find the motorcycle users tend to drive farther away or move to illegal-parking in response to high parking charge, and unlikely move to public transportation under single or loose restrained control strategy. Offering the space for motorcycle parking, implementing motorcycle parking charge, training traffic assistants for parking enforcement, and encouraging the private to open their parking spaces for the public are suggested for short term management. Augmenting parking control area and privatization of parking management are suggested for the medium term. In the long run, the roadside motorcycle parking should be prohibited everywhere.
