This study examines how leadership evaluation differs depending on the type of organizational crisis and the gender of leaders. Building on Coombs and Holladay's (1996) attribution-based typology, crises were categorized by locus of control and intentionality. This yields four types: internal-unintentional (e.g., production failures), internal-intentional (e.g., corporate scandals), external-unintentional (e.g., natural disasters), and external-intentional (e.g., terrorism). Collins et al. (2023) have noted that distinct leadership themes emerge across these types. Extending this framework, the present study incorporates the gender of leader. Bruckmüller and Branscombe (2010) suggested that "glass cliff" occurs because feminine characteristics (e.g., warmth) are suitable for leadership in times of crisis. So, we also examine how feminine characteristics (communion) and masculine ones (agency) of leaders are evaluated depending on the type of crisis. Online experiments were conducted in late 2024 with 800 participants recruited via a research company. Respondents were randomly assigned to one of eight conditions crossing 4 crises type (internal vs. external × intentional vs. unintentional) with leader gender (male vs. female). Each participant read a crisis scenario and a leader's short biography and then completed evaluation measures. Dependent variables included organizational reputation and leader evaluation (communion, agency, task-oriented leadership behaviour, person-oriented one and change potential). Two-way ANOVAs revealed that crisis type affected organizational reputation, and that leader gender affected person-oriented leadership behaviour significantly. Significant interactions emerged for task-oriented leadership behaviour, person-oriented one, and change potential. Notably, evaluations of male leaders remained stable across crises, whereas female leaders were judged differently depending on crisis type. These results echo Kokubo's (2023) findings that female leaders' evaluations vary with organizational performance, while male leaders' evaluations remain consistent. Overall, the study highlights the situational nature of leader evaluations, particularly for women, and suggests that gendered expectations shape perceptions of effective crisis leadership.