Women and members of other underrepresented groups who break through the glass ceiling often find themselves in precarious leadership positions, a phenomenon that has been termed the glass cliff. The glass cliff has been investigated in a range of domains using various methodologies, but evidence is mixed. In 3 meta-analyses, we examined (a) archival field studies testing whether members of underrepresented groups, compared with members of majority groups, are more likely to be appointed to leadership positions in times of crisis; (b) experimental studies testing whether members of underrepresented groups, compared with members of majority groups, are evaluated as more suitable for, as well as (c) more likely to be selected for, leadership positions in times of crisis. All 3 analyses provided some evidence in line with the glass cliff for women. Specifically, the meta-analysis of archival studies revealed a small glass cliff effect that was dependent on organizational domain. The leadership suitability meta-analysis also showed a small glass cliff effect in between-participants studies, but not in within-participants studies. The analysis of leadership selection revealed that women are more likely to be selected over men in times of crisis, and that this effect is larger in countries with higher gender inequality. The glass cliff also extended to members of underrepresented racial and ethnic groups. We explore several moderating factors and report analyses shedding light on the underlying causes of the glass cliff. We discuss implications of our findings as well as open questions.