A nonlinear control strategy-input/output linearizing by nonlinear state feedback-is suitable for trajectory tracking in batch processes. But based closely on the system model, the controller is sensitive to model parameter errors, which limit its application. In this paper, a technique of adaptive control of linearizable systems, together with an extended Kalman filter, was applied to a simulated batch styrene polymerization reactor to realize the output (extent of reaction) tracking in the presence of model parameter uncertainties. Simulation results show that this technique can ensure the output tracking performance even in the presence of two model parameter uncertainties, the first one being the gel effect the other one the heat transfer coefficient.