The aim of this study was to present a two-step optimization system to find the optimal process parameters of the multi-quality characteristics for transparent polypropylene (PP) composites with high impact resistance enhanced by rubber segment-styrene ethylene/butylene styrene (ERS-SEBS). Three selected proportions of PP/ERS-SEBS composites were made by single-screw mixing. Injection molding was applied to fabricate the PP/ERS-SEBS composites. In the first step of the optimization, Taguchi's orthogonal array was employed to arrange the experimental work. In the second step, the back-propagation neural network (BPNN) and particle swarm optimization (PSO) were applied to find the optimum parameter settings. The BPNN established the relationship between controllable parameters and the quality responses for the fitness function, which could be established for PSO. The BPNN quality predicator was combined with PSO to present the optimal parameter settings. The qualities, impact strength, tensile strength, and haze were investigated to find the optimal process parameter settings for the best quality specification. Control factors, such as the ERS-SEBS content, melt temperature, injection speed, packing time, and influence of cooling time on the process were discussed and optimized. Finally, the effectiveness of the proposed method was evaluated by the confirmation experiments. The impact strength and tensile strength of the PP/ERS-SEBS composite were raised significantly and maintained high transparency, thereby confirming that the combination of the BPNN and PSO would constitute the best process parameter settings.