Abstract According to the social identity theory, the group engagement model predicts discretionary cooperative behaviors are the consequences of organizational justice through a sense of identity security. The model states that “the central reason that people engage themselves in groups is because they use the feedback they receive from those groups to create and maintain their identities” (Tyler & Blader, 2003: 353). Examining the group engagement model, this thesis tests both procedural and interactional justice as the source brining about identity security, and measures the two components of social identity by organizational and work-unit identification, pride and respect. Based on a sample size of 124 dyads, the mediation effect is diverse for the two components of social identity that only the evaluative component partially mediates the impact of organizational justice on job satisfaction and voice behaviors. But it is alarming that none of the components could mediate justice on job performance.