A non-dominated genetic algorithm, called NSGA-II, is presented to solve the problem of allocating different firepower to shoot objects with different traits. Firstly, Mathematic model of artillery firepower allocating is analyzed. Then, details of NSGA-II are introduced, including encoding, sorting by non-domination, virtual fitness calculated, generating new individuals by backward operation, as well as the variation operation. Finally, the algorithm is analyzed by an application. The results show that NSGA-II can get optimal Pareto solutions effectively and efficiently, the outcomes are reasonable and truthful, it can provide a decision-making bases for commander.