By using the genetic algorithms and the strength formula
the optimal design of gear transmission can be carried out directly. The constraint conditions were transformed into binary code and fitness functions
so that the feasible region of optimization became into the range of the variables. In this way the mathematical model and the iterative procedure had been simplified. Three kinds of fitness functions based on different design requirements were proposed and practical disigning were carried out. Different solutions were obtained based on the different fitness functions so that the user can select the suitable solution according to the working condition of the gear transmission.