In recent years, the products have gradually developed towards portable products. The precision and the dimensional control of products are very important. Speciically, the warpage is one of the key factors to retain the quality of products. In order to overcome the problems encountered in complex product design and mold development more effectively, and to adjust the appropriate operating parameters, people apply computer-aided engineering technology (CAE) to assist product design and development. However, the results of CAE simulation are not consistent with that of real experiments. In this study we have utilized CAE simulation and experimental sudies to explore the differences bis etween simulation and experimental results, and to find out what the causes of the difference are. In Part I, we have applied Moldex3D software and real experiment on injection molding simulation based on the circle plate system with same process condition setting. Results show that the difference between simulation and experiment is 0.34 mm. To further find out what is the cause to make this difference, we have focused on the history curve o the injection pressure. Results show that the real injection time is delay about 29%, and the packing pressure is lower than about 23 %. To overcome this problem, we have setup one procedure to calibrate the real machine response based on the history curve o the injection pressure. Afer calibrated the machine response, the difference between simulation and experiment is reduced to 0.12 mm, and the correction rate is 64.7%. This calibration procedure has been verified using different material (from ABS to PP) and different geometrical design. All results show that the difference between simulation and experiment are reduced significantly. Moreover, in Part II of this study, we have focused on the integration of CAE and optimization techniques, specifically we applied Tuguchi method (called CAE-DOE) and Responxe Surface method (called CAE-RSM). In CAE-DOE study, before the machine calibration, the deviation from CAE-DOE simulation is 0.22 mm and that is 0.34 mm from real DOE experiment. However, after machine calibration, the deviation from CAE-DOE simulation is reduced to 0.05 mm and that is 0.17 mm from real DOE experiment. Obviously, the difference between CAE-DOE simulation and real DOE experiment is reduced sifgnificantly. Furthermore, In CAE-RSM study, before and after the machine calibration, the deviation of simulation can be reduced from 0.22 mm to 0.00 mm, and for the experiment it is reduced from 0.34 mm to 0.11 mm. Clearly, the machine calibration effect is quite important in optimization process in injection molding development.