Imprecise productions with low quality are produced by the incapable manufacturing processes. Prediction of the process capability in the design stage plays a key role to improve the product quality. In this paper, a new method is proposed to optimally reduce the locating error by allocating the minimum adjustments of locators. To quantify the precision of the manufacturing process, a proper tool that is called the geometric capability ratio (GCR) of the manufacturing process is introduced. First, based on a part fixture model, the relationship between the locating error and its sources is developed. Then, using the proposed geometric capability ratio, the manufacturing process capability is evaluated to achieve a specific desired level. If the process is incapable, the locating error should be essentially reduced. To improve the precision and accuracy of the final product, the error reduction procedure is developed as an optimal design problem. The formulated optimization problem can be efficiently solved by an evolutionary algorithm for constrained global optimization such as genetic algorithm method. The method is developed for the uncertainty analysis based on three approaches: the direct method, the worst case, and the statistical approaches. The proposed method is illustrated using a case study, and the computational results are compared to the obtained results from Monte Carlo and CAD simulations. © 2017, Springer-Verlag London Ltd.