Climate change is increasing the prevalence and impact of extreme events, which may have severe psychosocial aftereffects for the people and communities who are affected. To mitigate their impact, governments advocate developing community resilience. Most approaches to community resilience employ the concept of social capital, suggesting that communities with more dense preexisting networks of trust and reciprocity are more likely to prepare for, respond to, and recover more effectively from disasters. Notwithstanding the benefits of the social capital approach in explaining aspects of community resilience, I will argue that social capital cannot account for microprocesses of disaster behavior regarding the mobilization of communities and the emergence of groups in absence of pre-existing bonds or networks. Using empirical research I will demonstrate how community resilience theorizing can benefit through insights from the social identity approach in social psychology and principles of collective psychosocial resilience—the way that shared identification allows groups to emerge, coordinate, express solidarity, and provide social support. Overall, I will argue that this approach can overcome the limitations of social capital because it can explain the processes of group behavior in disasters, promotes bottom-up approaches to community resilience, recognizes emergent communities, and suggests evidence-based recommendations for policy and practice. The links between social capital and the social identity approach will also be discussed.