In this study, an optimal trajectory planning based on artificial neural networks and geneticalgorithms was proposed for control of nuclear research reactors. The trajectory being followed bythe reactor power is composed of three parts. In order to calculate periods of all parts of thetrajectory, a period generator was designed based on a feedforward neural network. Period values ofthe trajectory used to train the artificial neural network were acquired by utilizing genetic algorithms.The contribution of the proposed trajectory to the reactor control system was investigated.Furthermore, the behavior of the controller with the proposed trajectory was tested for various initialand desired power levels, as well as under disturbance. It was seen that the controller could controlthe system successfully under all conditions within the acceptable error tolerance.