Automated Guided Vehicle Systems (AGVSs) have been widely adopted by many low to medium manufacturing operations, particularly in Flexible Manufacturing Systems (FMS). The high degree of flexibility and control offered in vehicle routing has made AGVS a proven and viable material handling technology in today's manufacturing systems [Bozer91]. An important aspect in maintaining flexibility in an AGVS is its control architecture. A control architecture provides the backbone of the physical and the informational infrastructure of a system. This research has identified three types of control architectures. They are the centralized, hierarchical, and heterarchical control architectures. When designing an AGVS, most designers do not consider control architecture as a design factor, and do not analyze its effect on the system's performance. The objective of this research is to analyze the effect of control architectures on the relative performance of the AGVS. This research uses simulation to study the effect of control architectures on the AGVS. The simulation model for each control architecture contains two parts -- an AGV controller and a shop floor controller. Both models are programmed in C language. The AGV controller consists of three basic components – vehicle scheduling, vehicle routing, and traffic control. Each of these three components is modeled according to the nature and characteristics of the corresponding control architecture. Two different flow path layouts are considered for the shop floor model. The two layouts are different in size and number of work stations. Performance measures chosen for this study are intended to reflect the responsiveness of the system and the overall system performance under the impact of different control architectures. / Master of Science