Our climate is changing. Although the extent and rate of change are still uncertain, it is abundantly clear that the climate of the future will not resemble the climate of the past and will pose significant risks to people around the world. How well people adapt to address these risks will be determined by their adaptive capacity, their ability to instigate and implement change. Understanding and building adaptive capacity may therefore be key to reducing long-term vulnerability to global change. This dissertation clarifies the concept of adaptive capacity, synthesizes the substantial but largely unconnected body of research on adaptive capacity to date, and introduces a new methodological approach to conducting meta-analyses of adaptation science. I express a definition of adaptive capacity in mathematical terms that summarizes current theories on how adaptability is built, ties the concept to related concepts in adaptation, and poses questions about theoretical limits and thresholds for adaptation. I apply computational text analysis and network analysis tools to develop a concept model of adaptive capacity that identifies and organizes 158 determinants of adaptive capacity into 8 categories according to the functional role they play in building capacity. I propose a modular theory of adaptive capacity, in which all eight functional categories are critical but multiple pathways exist to achieve each function. This modular theory reconciles a theoretical debate in the literature and connects insights from existing theories with empirical findings from the field. I propose a new framework, the Adaptive Capacities Framework (ACF), based on the eight functional categories, that enables assessment of adaptive capacity across scales and within multi-scalar systems. Results demonstrate the fragmented nature of adaptive capacity research to date and propose new directions for future research. The dissertation also provides insights for practitioners seeking to prioritize adaptation efforts.