Abstract: The UCEP (uniform convergence of empirical probability) property that is a concept from statistical learning theory is combined with robust control. The design of the probabilistic robust stabilizing controller for control systems involving real parameter uncertainty is studied. Randomized algorithms based on the UCEP property are presented to solve probabilistic robust stabilizing controller design problems. Compared to traditional methods, randomized algorithms are more convenient to be used and can decrease the computational complexity dramatically. An example shows the efficiency of randomized algorithms.