Given the phenomenon of enormous observing over posting behaviors in virtual communities vis-à-vis relatively little exploration from academic, this paper investigates the importance of identity credibility under pre-interaction circumstance in virtual communities. We propose a model showing the antecedents and consequences of credibility of identity. Drawn on institution-based process, attitude change process and information richness theory, we identify three antecedents: perceived effectiveness of rules, formality of third-party endorsement which serve as central route in ELM, and information richness of target identity as peripheral route. Based on previous literatures, we also identify consequences from short term perspective (members' adoption intention of suggestions) and long term perspective (perceiver identity's likelihood of future direct interaction with the target identity) in the context of problem solving or knowledge sharing. We expect this model would provide virtual community practitioners with suggestions on how to improve credibility of their members through deliberate design and management of virtual communities.