In designing a high-speed motorized spindle-bearing system, bearing locations and tolerances must be designed optimally based on the concepts of motorized spindle shaft and bearings, and the number of bearings has been determined. Under certain constraints, a mathematical model of the optimization problem is developed to minimize total costs, which consist of the tolerance cost and expected value of the quality-loss cost. Therefore, we propose an optimization method that integrates the genetic algorithm and Monte Carlo Simulation to search simultaneously for the optimal values of nominal values and tolerance of location of bearings. To verify the effectiveness of the proposed method, this study investigates a real design example of bearing location for a high-speed motorized spindle-bearing system. Weights-related sensitivity analyses are also performed. The results of sensitivity study confirm that a higher tolerance cost results in a greater optimal tolerance design value that will be required.