Lately, researchers have paid increasing attention to negative or destructive side of supervisory behavior, i.e. abusive supervision. Studies on abusive supervision mainly focused on the results of abusive supervision. However, there are not many studies that examined the antecedents of abusive supervision. Recognizing this issue, our study examined the antecedents of abusive supervision in order to enhance our understanding of abusive supervision. Especially, we examined the effects of supervisor’s personality traits and supervisor’s perception of support from his/her manager on abusive supervision. Applying social exchange theory and trait activation theory, we developed and empirically tested the following hypotheses: H 1: Supervisor’s conscientiousness is negatively related to abusive supervision. H 2: Supervisor’s agreeableness is negatively related to abusive supervision. H 3: Supervisor’s perception of support from his/her manager is negatively related to his/her own abusive supervision. H 4a: Supervisor’s perception of support from his/her manager moderates the negative relationship between supervisor’s conscientiousness and abusive supervision such that the relationship is stronger when support perception is low than when support perception is high. H 4b: Supervisor’s perception of support from his/her manager moderates the negative relationship between supervisor’s agreeableness and abusive supervision such that the relationship is stronger when support perception is low than when support perception is high. Survey packages were distributed to 178 pairs of platoon leaders and their immediate subordinate, deputy platoon leaders and 160 were returned. Among those, 111 dyadic data were used for analysis. Platoon leaders’ questionnaires included their own conscientiousness, agreeableness, and support perception (i.e., company commander’s support) while deputy platoon leaders measured platoon leaders’ abusive supervision. The results indicated that supervisor’s agreeableness and perception of support from his/ her direct manger were negatively related to abusive supervision, as expected. Also, we found that the relationship between supervisor’s agreeableness and abusive supervision was stronger when supervisor perception of support from his/her manger was low. However, there was no significant relationship between supervisor’s conscientiousness and abusive supervision and no significant interaction between supervisor’s conscientiousness and supervisor perception of support from his/her manger. In summary, Hypotheses 2, 3, and 4b were supported. This research has theoretical and practical implications. First, our results suggested that supervisor’s agreeableness and his/her support perception played a significant role in determining abusive supervision. In addition, our findings revealed that abusive supervision was influenced by the interaction of supervisor’s agreeableness and support perception. These findings implied that although supervisor’s low agreeableness leads to hostile act toward his/her subordinates, it produces abusive supervision only among supervisors who perceive support in low. In other words, managers played critical role reducing the abusive supervision. This research is not free from limitations. First, the cross-sectional research design precluded any inference of causality. Second, data were obtained from the Republic of Korea army so the extent of generalizability of our findings to other organizations and other culture contexts remains an empirical question. Third, there was no significant relationship between supervisor’s conscientiousness and abusive supervision. Future research should repeat our study using narrow measure of conscientiousness (e.g., duty & achievement striving) in order to more fully understand the relationship between conscientiousness and abusive supervision. Finally, our study examined the limited variables as antecedents. Future research should examine various variables, including moderators, which are expected to influence abusive supervision.