Psychological contract(PC) is a key aspect in the employee-employer relationship,which develops over time in an organizational context.Few longitudinal studies exist regarding the nature of the employee PC and the causes of its change.A decade ago,three studies suggested that PC obligations would change in the first two years of employment(Robinson,Kraatz and Rousseau,1994;Robinson,1996; Thomas and Anderson,1998).Despite the fact that certain changes in employees' obligations to their employers have already been demonstrated,some important issues still remain to be addressed.The present study examined the change in the PCs of new entry-level college graduates in a Chinese organizational context.The study investigated how individual and contextual factors contributed to PC change over two time intervals.Specifically,the study examined the role played by individual preemployment beliefs concerning employment relations and organizational inducement in influencing changes in employee obligations.The following are the three major hypotheses proposed in this study:(1) initial changes in newcomer PCs will be in the direction of their existing beliefs;(2) organizational inducement including job rewards(JR) and organizational support(OS) will promote changes in employee obligations;specifically,JR is closely related to the transactional components of a PC,whereas OS is closely related to the relational components of a PC;(3) organizational inducement will moderate the relationship between the preemployment beliefs of new recruits and their PCs;specifically,JR will promote a positive relationship between short-term beliefs and changes in the transactional components of a PC change,whereas OS will enhance a positive relationship between long-term beliefs and the relational components of a PC. A longitudinal sample of college graduates was used to test the abovementioned hypotheses.Time 1 data were collected from graduates who had already secured full-time employment from various enterprises.Questionnaires were administered to 1,500 potential respondents in a leading domestic university;854 questionnaires were returned.One year later,the second wave survey was conducted on the 854 time 1 respondents;a total of 219 questionnaires were returned.By matching the two waves,the study employed 189 data points to test the hypotheses.Among the major measures,employee obligations were developed and validated in a Chinese organizational context,JRs were adopted from the Rusbult and Farrel(1983) study and the OS scale utilized was originally developed by Eisenberger et al.(1986) and later refined by Settoon et al.(1996).Cronbach's alphas for these measures were from 0.70 to 0.83,indicating acceptable measurement reliabilities. Hierarchical regression modeling(HRM) results indicated the following:(1) preemployment beliefs did not manifest direct effects on changes in employee obligations;(2) OS manifested positive impacts on both transactional and relational obligation change,and its impact on relational obligations was stronger than that on transactional obligations;and(3) the relationship between preemployment beliefs and changes in employee obligations was moderated by organizational inducement,which included OS and JR.Specifically,OS showed highly positive relationship with relational obligations when individuals held long-term beliefs.JR promoted both relational and transactional obligations of individuals who "did not" have long-term beliefs,whereas it had no impact on the obligations of individuals who had long-term beliefs. The above mentioned findings suggested the following.First,in comparison with economic exchange,relational exchange was more likely to be valued by Chinese employees.Hence,organizations operating in China should offer employees supportive working environments and emotional investment in order to achieve strong obligations from their employees and maintain the positive effect of long-term beliefs.Second,a moderate level of JR is necessary for organizations to maintain the obligations of employees with short-term beliefs.Limitations of the study and possible future investigation with regard to PC change were also discussed.