This paper presents a new strategy to optimize composite pressure vessels with metallic liners using an adaptive response surface method, Finite analysis is used to evaluate the fitness function and constraints, To reduce the computation time, the proposed algorithm creates the response surface (RS) model to find the optimum design. To improve the accuracy of the RS model, the design domain is iteratively reduced and new sample points are added to the RS model. In this way, the accuracy of the RS model is improved close to the optimum point. Thc genetic algorithm is used to find the reduced design domain and to generate new sample points. The optimization algorithm is first applied to a highly nonlinear function to evaluate its efficiency. It is then used to find the optimal design for a composite pressure vessel with a metallic liner.