Abstract This thesis applies the Genetic Algorithm to the solution of the optimal contract capacities for the customers in Taiwan Power Company System. From the actual automobile factory's data, it is quite obvious that the Genetic Algorithm can reach the optimal contract capacities very efficiently and make the electricity cost minimum in the whole year. To verify the effectiveness of the proposed approach, the thesis employs the real data from an automotive factory in Taiwan. The results obtained from the proposed approach are compared with that currently adopted by the factory for the power dispatching and capacity-contract setting. The numerical results reveal that about seven hundred and twenty thousand NT$ of electrical power expenses can be saved. Keywords: Capacity contract, Genetic algorithm