This paper focuses on the pickup and delivery location problem in automated guided vehicle (AGV) systems, which is the problem of locating pickup and delivery points of each workstation with the objective of minimizing the total distance of material flows between the pickup and delivery points. Here, the material flow distance between a pickup and a delivery points is defined as the product of the length of the shortest path and flow intensity between the pickup and delivery points. In this paper, we first suggest a procedure for eliminating dominated positions for pickup and delivery points. Then, to find good solutions within the solution space reduced through the elimination procedure, we develop three simulated annealing algorithms. Performance of the suggested algorithms is tested on randomly generated test problems, and results show that the suggested algorithms perform better than an existing method. Also, it is found from the computational results that the elimination procedure is very efficient and helps to obtain good solutions within a reasonable amount of computation time even for very large sized problems.