Abstract: In the optimization design of the optical film system, the evaluation function based on film system structure is a complicated multi-peak function. Traditional optimization algorithms depend heavily on the initial design of the film systems, and are prone to fall into the local extreme values close to the initial structure. To solve the problem, the global-optimized adaptive evolutional algorithm is applied to the film-system optimization design of optical thin film. In the design process, only the solution code and the fitness function are needed and no additional information such as initial design is required. The evolutional algorithm was compared with the popular film system design methods. The optimization examples of high reflective films and splitting films show that the algorithm can be used to obtain more reasonable film system structure under the same design requirement. The design is flexible and films can be easily prepared.