Abstract: Multi-objective bi-level optimization model for the control system of gun-launched missile is discussed. A bi-level hybrid genetic optimization algorithm for multi-objective is presented to solve the control of gun-launched missiles. First, by introduceing a Lagrange multiplier vector, the first-level model is optimized based on a hierarchical optimal. Then, combining the genetic algorithm to the pattern search, the second-level model is optimized. Simulation experiments showed that this hybrid genetic optimization algorithm, when compared to the known genetic algorithm, had faster convergence and the optimized control system of the gun-launched missile can track the maneuvered acceleration more accurately and rapidly.