This study examined the effects of leaders’ gender and the gender role of the leaders’ behavior on the leaders’ evaluations and the possibility that these effects were mediated by attributions to the leaders’ behavior. The study conducted two experiments. Experiment 1 (n=151) indicated that leaders’ evaluations were higher when male leaders acted consistently with women’s behavior. Experiment 2 (n=110) showed that female leaders’ evaluations were higher when female leaders acted consistently with men. These findings indicated that the leaders’ ratings were higher when their gender and behavioral gender roles did not match. Moreover, Experiment 2 showed that attributing leaders’ behaviors to the leaders’ gender partially mediated the effect of gender-behavior mismatch on leaders’ evaluation. The author has discussed the problems related to these findings and the prospects for this research.