In an effort to control costs, airlines have begun to concentrate on their maintenance operations as a potential source for savings. Nevertheless, federal regulations and internal safety policies effectively limit cost savings to improvements in productivity and scheduling. The purpose of this paper is to present a model that can be used by planners to both locate maintenance stations and to develop flight schedules that better meet the cyclical demand for maintenance. The problem is formulated as a min-cost, multicommodity flow network with integral constraints, and solved using a two-phase heuristic. The procedure is demonstrated with data supplied by American Airlines for their Boeing 727 fleet. The results show a significant improvement over current techniques, and indicate that substantial cost reductions can be achieved by eliminating up to 5 of the 22 maintenance bases now in operation. Similar results were obtained for American's Super 80 and DC-10 fleets. Perturbation analysis confirms the robustness of these findings, and suggests that loss in flexibility due to interruptions in the flight schedule will be negligible.