Abstract The analysis of an assembly process in the discrete part manufacturing industry usually involves a large number of dimensions. Each dimension tolerance plays a different role with respect to the dimension variation of the final product, which is the so-called “quality”(dimension precision). Furthermore, there are many random factors (noises) present during the assembly operations (tool wear, loose fixture, etc.). Few mathematical models can represent the assembly process. Therefore, computer simulation has been employed. Variation Simulation Modeling uses the Monte Carlo sampling technique to simulate the assembly process. By applying a geometric standard and simulating the physical operations, the statistics of the final product dimensions can be predicted. With the simulation results, statistical analysis is essential to identifying the critical factors (component dimensions). The traditional experimental designs, such as full factorial design, however, are not practical since the number of factors is too large. Taguchi method, which explores a special subset of factor combinations (called the orthogonal array) is able to examine a large number of factors (and interactions) in a much smaller number of experiments. The analysis of variance is performed to ensure the proper selection of significant factors. With this proposed unified tool, engineering understanding and judgement become more effective in making appropriate decisions regarding the product and process designs.