Application of optimization based techniques in control and operational optimization schemes of semi-batch batch reaction processes allow the rigorous inclusion of economic as well as safety aspects. For a restricted class of reaction processes, the on-line optimization problem can be reduced to optimization on a short prediction horizon requiring only crude process models. This paper extends an earlier approach of Helbig et al. (1998) of simultaneous temperature control and feed rate optimization in the framework of model predictive control (MPC). Based on simple calorimetric process models, the given MPC formulation is readily applicable in typical industrial multi-product or multi-purpose plants. The major capability of driving the process along constraints is illustrated in a simulation study of an isothermal semi-batch reaction process with a single reaction.