Abstract:  Carsharing service, as a new selfservice travel mode, has joined the public transportation field In order to study the characteristic variables that affect potential users′ selfefficacy, data from 318 potential users′ selfefficacy in carsharing services in Beijing, Shanghai and Guangzhou is collected by Ipsos Group According to hypotheses, the differences among overall selfefficacy, vehicle selfefficacy and application selfefficacy in different variables (gender, age, city, income, bicyclesharing service and driving license) are discussed and analyzed in detailResults show that gender, age, city, income and driving license are all significant predictors of potential users′ selfefficacy However, whether participants use bicyclesharing services or not has no significant effect on selfefficacy in carsharing servicesResults provide important theoretical value and practical guidance for the carsharing services  marketing

Key words: carsharing services, selfefficacy, empirical research