Building evacuation planning is done with the goal of evaluating the egress of building occupants including the exit locations, the escape routes, and the corresponding egress times. The traditional approach utilizes drills which usually incur corresponding costs and inconvenience. In this study, a 3DGIS-based geosimulation model for building evacuation is developed using the GAMA simulation platform as an alternative and complement to actual evacuation drills. PEOPLE agents are modeled to represent the building occupants. FLOORS, ROOMS, INDOOR_PATHS and EXIT_POINTS agents are modeled from the 3DGIS building layers of the case study building. The simulation involves PEOPLE agents evacuating to their target EXIT_POINTS along the INDOOR_PATH. Their movement is affected by their chance_panic (CoP) and chance_knowledge_exit (CoKE) attributes. If the CoP value is true, the speed of the PEOPLE agent is multiplied by an assumed value of 1.5. Otherwise, they retain the base speed of 0.83 m/s. If the CoKE value is true, the PEOPLE agent is assigned the target EXIT_POINT corresponding to the designated_exit attribute of the ROOMS where the PEOPLE agent is located. If CoKE is false, the PEOPLE agent selects a random exit. Weights applied to the INDOOR_PATH computed from the number of PEOPLE agents currently on the same INDOOR_PATH also affect the movement speed. The egress time periods for different simulation runs were computed and analyzed to examine the effects of varying CoP and CoKE values on the evacuation time. In general, the increase in CoP values resulted to an increase in the total egress time. Extreme values were obtained at CoP equal to 1 in which the movement speed of the PEOPLE agents greatly decreased the total egress time. With CoP equal to 1, the increase in CoKE still caused decrease in the total egress time. As CoKE is increased, the total egress time decreased correspondingly. At CoKE equal to 1, the increase in CoP still contributed to increased total egress time. From these results, it can be concluded that even if panicking PEOPLE are moving faster, egress time is reduced if they have knowledge of the exit. Likewise, even if PEOPLE have knowledge of the exits, they can evacuate more efficiently if they do not panic.