The beneficial role of political skill in stress reactions and performance evaluations has been demonstrated in a substantial amount of empirical research. Most of the research, however, has focused on self-perceptions of political skill. This study examines the differential moderating effects of self- vs. other-rated political skill in the conflict--emotional burnout and performance relationships, using two samples including non-academic staff employees of a large university (N = 839) and a variety of office and retail employees from an automotive organization (N = 142). We argue that self-reported political skill moderates the relationship between conflict and a self-reported strain-related outcome that is important to the individual (i.e., emotional burnout), but that supervisor-rated political skill does not moderate this relationship. Further, we argue that supervisor-rated political skill moderates the relationship between conflict and an outcome important to the supervisor and the organization (i.e., job performance), but that self-reported political skill does not moderate this relationship. Findings partially support our hypotheses as both self and supervisor-rated political skill neutralized the negative effects of conflict on burnout, but only supervisor-rated political skill neutralized the negative effects of conflict on performance. Limitations and directions for future research are discussed. (Contains 3 figures and 4 tables.)