Modeling, optimization and control techniques are indicated to improve the productivity, yield and efficiency of biotechnological processes. The availability of on-line measurements of the state variables plays an important role in the application of these techniques. The most common measurements are pH, temperature, dissolved oxygen, CO2, among others. However, it is also necessary to measure substrate and product concentrations efficiently. Biosensors can be used for that purpose. This paper investigates the use of biosensors-coupled with 'flow injection analysis' (FIA)-for process control. The experimental use of an enzymatic lactose biosensor in a real fermentation media is described. The nonlinear nature of the biosensor is illustrated and modeling approaches of these systems are introduced with the objective of providing tools for the optimization of the process. (C) 2003 Elsevier Science Ltd. All rights reserved.