This paper deals with an application of Intelligent Computer Aided Learning in the field of control engineering. First we analyze Conventional Computer Aided Learning, with th 3 use of quantitative Computer Aided Control System Design packages. The limitations of these systems for learning are explained. This paper then presents an environment relying on symbolic knowledge, connected with a numerical simulator, dedicated to train the students qualitative interpretation of physical process behavior. A qualitative behavior classification is proposed, linking symbolic terms (such as stable, oscillatory, over‐damped,...) to parameter ranges in the differential equations describing the process. Then, the environment interaction, its knowledge organization, and its cooperative function are detailed, together with its object oriented modeling.