We propose in this paper a genetic algorithm (GA)-based approach to solve the Flexible Manufacturing System (FMS) formation problem. First, we divide the unity of machines into small groups or cells using Group Technology (GT). And an option for considering the minimization of workload variation between cells is included. Then the implementation is done using GA. When using GA, the designer is allowed to specify the number of cells and impose lower and upper bounds on cell size. This makes the GA scheme flexible for solving the FMS formation problems. We propose in this paper a genetic algorithm (GA) -based approach to solve the Flexible Manufacturing System (FMS) formation problem. First, we divide the unity of machines into small groups or cells using Group Technology (GT). And an option for considering the minimization of workload variation between cells is included. Then the implementation is done using GA. The using GA, the designer is allowed to specify the number of cells and impose lower and upper bounds on cell size. This makes the GA scheme flexible for solving the FMS formation problems.