There is a growing need to increase the performance of the built environment through a combination of improved design, retrofitting of existing structures, and behavioral and policy change. Increased performance includes decreasing construction and operational costs, improving efficiency, reducing energy consumption and overall carbon footprint, and increasing the health, safety, and comfort of building occupants. Data collection and analysis are central to ongoing efforts in performance improvement. The growth of sensor and monitoring technologies, coupled with the proliferation of building automation systems, is quickly leading to an explosion in the amount, quality, and format of building performance data. What is needed are methodologies for extracting viable information from these data and using the results to effect meaningful change. Furthermore, occupant behavior and attitudes must be taken into account. This paper summarizes agent-based simulation and describes its potential as an approach to support analysis, design, and performance improvements in the built environment.