An adaptive fuzzy Smith control method is proposed in this paper to control time-varying processes with dominant and variable time delay. By minimizing the error between the process output and the process model output, the variable delay parameter is identified. Based on the on-line parameter estimation, a new Smith predictor is obtained which can compensate for the effect of time delay completely. This method can alleviate the inherent drawback of Smith predictor that its control performance is sensitive to the process model. Furthermore, an adaptive fuzzy controller which can adjust on-line the scaling factors is also incorporated to improve the control performance. Simulation results have demonstrated the effectiveness of the proposed approach.