The configurations of reconfigurable manufacturing systems (RMS) change with time to provide functionality and capacity needed for every demand period. This paper presents a novel model for optimizing demand period cost (fixed plus operating) of RMS flow-line configurations. The flow-line configurations being considered include arrangement of machines (number of stages and number of parallel machines per stage), equipment selection and operations assignment. Input requirements consist of an operation sequence, machine options for each operation in this sequence and operation capabilities for all available machines. Generating K economical configurations subjected to constraints on space, initial investment, functionality and capacity is modeled as a constrained K-shortest paths problem on an augmented acyclic digraph. Subsequently, a graph theory based approach is introduced to find the optimal and K-1 suboptimal configurations. Finally, a case study shows that the proposed method is valid and efficient.