We present a discrete-time, Mixed Integer Linear Programming (MILP) model for the production scheduling of a continuous-process multi-grade PET resin plant. The objective is to minimize the cost associated with grade changeovers in order to avoid undesirable variations in base resin properties and process conditions that occur during such changes. The constraints of the model include requirements related to sequence-dependent changeovers, sequential processing with production and space capacity, mixed and flexible finite intermediate storage, and intermediate demand due-dates. We present a case study that illustrates the application of the model on a real problem scenario and provides insight into its behavior. The numerical experience demonstrates that the computational requirements of the model are quite reasonable for problem sizes that typically arise in practical applications. (C) 2009 Elsevier Ltd. All rights reserved.