An individual’s perceived risk is an important source for determining responses during a fire evacuation. The level of perceived risk is continuously influenced by an interpretation of the updated emergency information from the surroundings. In this regard, this research aims to model and analyze the effect of perceived risk that changes from observing dynamic emergency cues (i.e., information from the fire itself, other occupants, and facility emergency systems) on evacuation behaviors in public building fire situations using an agent-based modeling approach. The simulation results highlight the difference in the effect of dynamic emergency cues on evacuation performances based on the type of emergency cues and facility layout. In addition, the result emphasizes the necessity of considering the side effects of dynamic emergency cues on the evacuation performance. Using the developed model, this research provides insights into the appropriate use of dynamic emergency information considering the extended influence of perceived risk on evacuation performance.