This paper addresses the reconfigurable layout problem, which differs from traditional, robust and dynamic layout problems mainly in two aspects: first, it assumes that production data are available only for the current and upcoming production period. Second, it considers queuing performance measures such as work in progress inventory and product lead time in the objective function of the layout problem. A process to solve the reconfigurable layout problem is proposed. A previously developed, open queuing network-based analytical model, called Manufacturing system Performance Analyser, is used to estimate the stochastic performance measures of a layout. These are combined with deterministic performance measures such as material handling cost to determine the layout for the next period. A case study is used to illustrate the process of solving the reconfigurable layout problem.