The Reconfigurable Manufacturing Systems (RMSs) paradigm provides systems with a high degree of capacity, scalability and functional adaptability through configurations. Our discussion deals with Multicriteria Decision Making (MCDM) within an RMS context to facilitate choice in case of indeterminism, and to show that the “near-most appropriate” configuration is selected when unexpected events occur. More precisely, it focuses on the use case of MCDM to carry out the request in the current configuration, and to classify the set of configurations which are able to carry out this request, taking into account a hierarchical criteria system in both of these cases. In other words, product scheduling and the configuration are two closely related issues which have a strong impact on each other. Therefore, we propose the Analytic Hierarchy Process (AHP) to obtain near-optimal solutions and get an evaluation method to classify configurations that can be used for reconfigurable manufacturing systems when they are faced with sudden failures.