This paper reviews studies concerned with abusive supervision and provides a constructive revision of Tepper's 2007 model. As a result of our review of the recent research, we revised the 2007 Tepper model and added additional variables and casual paths to increase its explanatory potential. The model we propose distinguishes between abusive supervisory behavior and abusive supervisory perceptions, suggesting that each of these variables needs to be studied separately until we know more about how they are related. The revised model also explicitly recognizes possibilities for reverse causation and stresses the importance of subordinates' individual differences such as attribution style, negative affectivity, and implicit work theories, which have the potential to account for significant variability in subordinates' perceptions of abuse. Suggestions for future research based on the original relationships identified by the Tepper review as well as the variables and causal paths suggested in the revised model are provided. emerge as the result of our review. Next, we propose a constructive revision of the Tepper model, which addresses some of the limitations of the prior model and provides suggestions regarding the causal paths and variables that have the potential to explain significant variance in the outcomes associated with abusive supervision. We end with suggestions for future research.Before conducting our review of the recent research, we examined Tepper's (2007) review and the articles covered in his summary to make sure we understood his perspective. Although we agreed with many of his observations and recommendations, we also had some concerns about this body of research. These concerns centered around six issues. First, as noted by Tepper (2007), a majority of the data in these early studies was from single sources (usually the subordinate), and almost all of the research designs were cross-sectional. As a result, we are concerned that strong causal inferences are being made despite the limitations of these designs.Second, we are concerned that researchers appear to be assuming that commonly used abusive supervision measures are objective and reliable measures of abusive supervisory behaviors. Thus, for example, when discussing the results of his research on abusive supervision, Tepper (2000, p. 186) stated "subordinates whose supervisors were more abusive reported higher turnover, less favorable attitudes toward job, life, and organization." This would not be a problem if the statement said "subordinates whose supervisors were perceived to be more abusive" as opposed to "were more abusive," as the Tepper scale measures perceptions rather than behaviors. Third, we are concerned by the lack of validation of abusive supervision measures. We could not find any objective measure of abusive behavior that was related to Tepper's (2000) commonly used perceptual measure.Fourth, we are also concerned that researchers are ignoring plausible theoretical alternatives to the origins of perceptions of abusive supervis...