Here we Show how neural networks can be used to learn how to schedule job shop problems through learning and generalizing the expertise implicitly from several job shop schedules. A work based on training a multi-layered perceptron (MLP) network to solve the job shop scheduling problem is presented. The shortest processing time SPT rule is used for obtaining feasible schedules for 6×6 job shop problems and the neural network is trained on these schedules. MLP neural network was trained using the back propagation “BP” learning algorithm. Finally feasible schedules have been developed according to neural network model output and a developed algorithm. The performance of the proposed model is compared to the performance of branch and bound, extracted rules and simple dispatching rule SPT in scheduling a test set of scheduling instances and 10×10 scheduling instance.