In this study, we constructed an automatic optimization system applying the multiobjective genetic algorithm (MOGA) and developed an artificial implantable lung possessing antithrombogenicity and high gas exchange performance based upon fluid dynamics. This system consists of a three dimenstional CAD system, computational fluid dynamics software, and the multiobjective optimization tool modeFRONTIER (ESTECO CO., Trieste, Italy). The objectives were to minimize the volume of the region having a flow rate of less than 0.5 mm/s by assuming that thrombus formation occurs at this limit (ObjTF) and to minimize the standard deviation of the flow rate in the hollow fiber to obtain high gas exchange performance (ObjGEP). In optimization 1, the arc heights (six variables) and the distance between cross-sections (two variables) were used as design variables in the inflow and outflow portions. In optimization 2, the edges (two variables) of the inflow and outflow portions were optimized in the resulting designs from optimization 1. The optimum designs were manufactured using the rapid prototyping system and were examined by evaluating gas exchange performance (ObjGEP) in vitro. Gas exchange performance increased as the improvement ratio of ObjGEP became higher. For the optimum design (improvement ratio of 74.8% for ObjGEP), O2 transfer increased by an average of 18.4%, and CO2 transfer increased by an average of 40.5% when compared with the original design. The results suggest that this system was not only effective for reducing the time, cost, and labor of developing artificial organs but was also useful as a design and development support system for high performance artificial organs for transplantation.