The FireGrid project aims to harness the potential of advanced forms of computation to support the response to large-scale emergencies (with an initial focus on the response to fires in the built environment). Computational models of physical phenomena are developed, and then deployed and computed on High Performance Computing resources to infer incident conditions by assimilating live sensor data from an emergency in real time—or, in the case of predictive models, faster-than-real time. The results of these models are then interpreted by a knowledge-based reasoning scheme to provide decision support information in appropriate terms for the emergency responder. These models are accessed over a Grid from an agent-based system, of which the human responders form an integral part. This paper proposes a novel FireGrid architecture, and describes the rationale behind this architecture and the research results of its application to a large-scale fire experiment. ► Demonstration of infrastructure for urgent emergency response decision support. ► A simulation model infers incident state that is interpreted by knowledge reasoning. ► Dense sensor networks provide live data for steering simulations in real time. ► The integration of Grid and HPC provides requisite computational power. ► AI techniques rationalize and present complex simulation results in a concise manner.