Drawing from moral exclusion theory, this article examines outcome dependence and interpersonal liking as key boundary conditions for the linkage between perceived subordinate performance and abusive supervision. Moreover, it investigates the role of abusive supervision for subordinates' subsequent, objective work performance. Across 2 independent studies, an experimental scenario study (N = 157; Study 1) and a time-lagged field study (N = 169; Study 2), the negative relationship between perceived subordinate performance and abusive supervision was found to hinge on a supervisor's outcome dependence on subordinates but not on a supervisor's liking of subordinates. Furthermore, Study 2 demonstrated (a) a negative association between abusive supervision and subordinates' subsequent objective performance and (b) a conditional indirect effect of perceived performance on subsequent objective performance, through abusive supervision, contingent on the degree of outcome dependence, although these relationships did not reach conventional significance levels when controlling for prior objective performance. All in all, the findings highlight the role of instrumentality considerations in relation to abusive supervision and promote new knowledge on both origins and consequences of such supervisory behavior.