Summary: In biotechnology, as is many other fields of technology, the development of a suitable process model is one of the most important engineering tasks. Data driven modelling becomes an attractive approach whenever analytical modelling is difficult or too time consuming. Disadvantages of the often used artificial neural networks are their missing transparency and the difficulty to integrate prior knowledge. The paper at hand gives an overview of several common modelling techniques with focus on their application to bioprocesses and presents a novel modelling technique that uses genetic programming for the construction and refinement of transparent structured models.