Abstract: Based on the pole on-line optimization principle and stochastic process theory, a parameter adaptive and quadratic performance optimal combined control for slowly time-varying stochastic servo systems is proposed. With simple form and modified computational effort, this approach can fairly adapt to the time-varying characteristic of the systems and maintain system's optimal tracking property under different parameters and various working conditions. In the paper, convergence of the control is. also discussed and the effectiveness and robustness of the control is shown by simulation and practical application.