Recent studies of organizational behavior have witnessed a growing interest in unethical leadership, leading to the development of abusive supervision research. Given the increasing interest in the causes of abusive supervision, this study proposes an organizing framework for its antecedents and tests it using meta analysis. Based on an analysis of effect sizes drawn from 74 studies, comprising 30,063 participants, the relationship between abusive supervision and different antecedent categories are examined. The results generally support expected relationships across the four categories of abusive antecedents, including: supervisor related antecedents, organization related antecedents, subordinate related antecedents, and demographic characteristics of both supervisors and subordinates. In addition, possible moderators that can also influence the relationships between abusive supervision and its antecedents are also examined. The significance and implications of different level factors in explaining abusive supervision are discussed.