Organizational change and innovation are critical for business survival and more likely to occur when employees engage in change-oriented behaviors. Previous studies have examined the direct effects of workplace social exchanges on employees’ change-oriented behaviors; however, less attention has been given to the combined effects of these exchanges and the mechanisms by which these relationships occur. In this study, I look at the combined effects of leader-member exchange, trust in team members, and perceived organizational support on voice, innovative, and learning behaviors via psychological safety. In addition, based on the understanding that psychological safety is not always present in the work environment, I look at the conditions under which these workplace social exchanges lead to the aforementioned behaviors even when psychological safety is low or absent. Therefore, the purpose of this dissertation is three-fold: (1) to explore the combined effects of leader-member exchange, trust in team members, and perceived organizational support in improving followers’ psychological safety within the organization, (2) to investigate the mediating role of psychological safety on the relationships between workplace social exchanges and followers’ change-oriented behaviors, and (3) to expand on previous findings and examine the conditions under which these social exchanges and psychological safety lead to followers’ change-oriented behaviors. Specifically, I propose and test a theoretical model derived from social exchange theory to examine conditional indirect effects of leader-member exchange, trust in team members, and perceived organizational support on voice, innovative, and learning behaviors through psychological safety within the organization, and to examine the role of proactive personality, political skill, perceived team social integration, perceived support for innovation and perceived organizational justice as second stage moderating variables that may compensate for low psychological safety within the organization. My theoretical model was tested using lagged data collected from leader-follower dyads representing 174 followers and 85 leaders from four organizations located in the United States. To test this theoretical model, I used a quantitative non-experimental research design, a survey method, and multilevel analytical procedures.