We describe a new architecture to integrate a psychological model into a crowd simulation system in order to obtain believable emergent behaviors. Our existing crowd simulation system (MACES) performs high level wayfinding to explore unknown environments and obtain a cognitive map for navigation purposes, in addition to dealing with low level motion within each room based on social forces. Communication and roles are added to achieve individualistic behaviors and a realistic way to spread information about the environment. To expand the range of realistic human behaviors, we use a system (PMFserv) that implements human behavior models from a range of ability, stress, emotion, decision theoretic and motivation sources. An architecture is proposed that combines and integrates MACES and PMFserv to add validated agent behaviors to crowd simulations.