In this work, the injection molding process of an automotive part made of PA 66 is optimized virtually by using simulation and Design of Experiments. A mold model including hot runner and cooling system is built in the simulation software Cadmould 3D-F 6.0 and the serial production process is simulated. The average temperature over the whole part thickness, the mold surface temperature distribution, the sealing times, the shrinkage and the warpage, respectively, will be considered in more detail. By means of the process optimization software VARIMOS 4.06 selected parameters (filling time, holding time, holding pressure profile, cooling time, cooling fluid temperatures, melt and hot runner temperature) are systematically varied in an D-optimal experimental design. Based on the simulation results, linear mathematical process models for part width, part height and warpage will be calculated to determine both, the major influence parameters and an optimum robust process setting, where all quality parameters are within their given tolerances. Furthermore, the use of mold materials with a higher thermal conductivity and the disturbance parameter "calcium deposit in the cooling channels" are simulated to analyze cycle time dominating hot spots. Finally, the feasibility of PBT GF20 as alternative polymer using both, the given part design and the given mold is proved. In several simulations based on a D-optimal experimental design an optimum process setting is determined.