As the emissions regulations have become more stringent, reducing NOX emissions is of great importance to the shipping industry. Due to the price and emissions advantages of natural gas, the diesel-natural gas engines have become an attractive solution for engine manufacturers. Firstly, in this paper, the NOX emissions prediction model of a large marine four-stroke dual-fuel engine is built by using AVL-BOOST. In addition, the model is further calibrated to calculate the performance and emissions of the engine. Then, the influences of boost pressure, compression ratio, and the timing of intake valve closing on engine performance and emissions are analyzed. Finally, the response surface methodology is used to optimize the emissions and performance to obtain the optimal setting parameters of the engine. The results indicate that the response surface method is a highly desirable optimization method, which can save a lot of repeated research. Compared with the results from manufactured data, the power is increased by 0.55% and the BSFC, the NOX emissions, and the peak combustion pressure are decreased by 0.60%, 13.21%, and 1.51%, respectively, at low load.