Under the stochastic demand, one of the key problems for RMS design is how to design the capacity scalability and storage volume. Based on queuing theory, stochastic market demand model for single product family and production model of manufacturing systems were built. The queuing model and the state probability transition graph of RMS were given. The probability balance equations under stationary condition were constructed. Performance measure indexes of RMS and FMS were presented. Finally, a case study illustrated the developed method is valid. The performances between RMS and FMS were compared, and the advantages of RMS were shown.