Performance-based design has become inherently important in construction. It is absolutely necessary to assess the level of safety provided by the building structure to its occupants in normal conditions and more importantly in fire incidents, since code compliant structures don’t always offer the needed safety for occupants. This can be achieved by combining static data from building codes and regulations with the dynamic data resulting from the interaction between occupants and between occupants and their surrounding environment. This research proposes a framework that uses building information modelling (BIM) and agent-based simulation (ABS) to assess the evacuation performance of buildings. BIM is used to model the geometry of buildings and to visualize the evacuation process. It acts as a platform for ABS to identify potential bottlenecks and points of hazard, based on the data collected regarding agents behavior in emergency situations. In order to utilize the simulation results in an effective manner, they are integrated into a decision support system (DSS) to help in making better decisions. That’s why the simulation results as well as other data are fed to the decision making model. The developed framework is capable to rank and assess different evacuation plans and/or design alternatives depending on several criteria, unlike other systems that evaluate buildings safety using the egress time as the sole criterion. A case study is presented to illustrate the practicality of the proposed system. The building’s spatial characteristics as well as the occupants’ data (e.g., approximate numbers, percentages of the disabled, and behavioral data such as: movement speeds and pre-evacuation times) are identified to study their potential impact on the evacuation. A number of recommendations, concluded from the conducted research and case study, are presented at the end of the paper.