With increasing demands of machining accuracy, designing of machine tools for satisfactory performance using cost-effective geometric accuracy configurations is becoming a complex problem to the machine tool manufacturers. In this paper, a novel robust accuracy allocation method is proposed for multi-axis machine tools based on multi-objective quality and cost trade-offs. To model the volumetric accuracy of machine tool based on geometric errors, the multi-body system theory was introduced. A manufacturing cost model for the machine tool components with a significant effect on geometric errors was established based on the machining features. The quality loss of the machine tool was also integrated into a single optimization objective. After identifying the relationship between the accuracy grade parameters of the feeding components and the geometric errors, the maximum in the Euclidean norm of all the accuracy parameters was defined as another optimization objective. The robust accuracy allocation was performed using Isight software and the Non-Dominated Sorting Genetic Algorithm-II built in the MATLAB. The optimization results for a four-axis horizontal machining center showed that the proposed method can realize the optimization of geometric accuracy and can determine the optimal accuracy grade of the feeding components satisfying the machining accuracy requirements.