Automatic controls are an essential part of every modern IC-engine. Usually these controllers are linear with scheduled gains and require a large amount of engine-testing to work satisfactorily in all operating conditions. Nonlinear and adaptive controllers - whose design is based on physical models of the engine - have the potential of improving the performance of the engine and to reduce the required testing time. This paper shows how nonlinear models and controllers fit into a general engine-control frame-work. The presented ideas are clarified by re-analyzing the well-known Air/Fuel-ratio control problem from a nonlinear point of view.