This paper presents the application of model based predictive techniques to an olive oil mill. The work presents a solution to the integrated control of the mill, where a predictive strategy has been used to optimize oil yield while keeping quality standards. The work includes multivariable identification as well as the implementation of a predictive controller on the real plant, taking into account operating constraints that appear in this process. The work also shows the application problems that arise when implementing advanced controllers in an industrial control system with low computational capabilities. The application of the proposed control strategy to an actual olive oil plant has shown that great benefits can be obtained both in oil yield and extraction performance. (C) 2007 Elsevier Ltd. All rights reserved.