In recent years, the continuous growth of private cars, the tight supply of land resources and the poor air quality have led policy makers to advocate sustainable public transportation. Bike sharing system has been introduced by many cities and developed worldwide rapidly, due to its advantages in reducing environmental pollution and alleviating traffic congestion. It is recognized as a strategic tool to integrate public transportation and promote sustainable urban transportation. Cities around the world seek to reshape urban transportation to a greener and healthier way with the help of bike sharing system. However, with the development of bike sharing system, many challenges related to its operation have not been well solved, including modeling the demand and encouraging users to purchase memberships. This thesis focuses on modeling the demand of bike sharing system and exploring the factors that affect users’ subscription membership, in the context of Citi Bike initiative in New York City. An online survey is designed to collect data of both characteristics and hypothetical decision of a discrete choice experiment. Discrete choice models are built to give insights into users’ choices.