Abstract: A mathematical model to optimize the preloading of large net-shape deployable antennas was established , with the magnitude of the preloading on vertical cables as the design variable, the root mean square on the net-shape reflective surface as the objective function, and the stress and the frequency as constraints. Optimization using Genetic Algorithm was then performed. As a result, the proposed method can effectively provide a group of preloading, thus obtaining high surface accuracy of the initial shape of the net-shape antenna. The experimental results may serve as a good reference for engineering designers.