The main objective is to improve the control performance during catalyst processing experiments in a furnace heated fluidised bed reactor. A model-based predictive control strategy is implemented for the tight control of the reactor temperature and the overall energy requirements. A detailed dynamic model is developed for the furnace reactor and validated using open-loop experimental data. Dynamic programming techniques are used for the calculation of the optimal sequence of manipulated variables that minimise the difference between the desired and the predicted reactor temperature trajectories. The model predictive control strategy leads to superior performance than the previously installed system that consisted of conventional PID controllers. (C) 2005 Elsevier Ltd. All rights reserved.