This paper is concerned with the optimization of biological production processes. The process is characterized by a conflict between yield and productivity as revealed by the analysis of the model. The optimization problem is to find an operating mode that achieves the best trade off between yield and productivity. It is shown how this problem can be formulated as a parametric optimization involving parameters that have a clear engineering meaning. This parametric optimization leads to the definition of a control problem which requires a feedback implementation under the form of an adaptive regulator combined with a software sensor. A simple mechanistic model is considered as a benchmark example for this optimization study. (C) 1998 Elsevier Science B.V. All rights reserved.