Abstract: Based on the analysis of a certain amount of on-site test samples, this paper constructs a high-precision artificial neural network model for the ductile-brittle transition temperature prediction of RPV materials. Then we use the model to explore the influence of neutron fluence and neutron fluence rate parameters on the ductile-brittle transition temperature of RPV materials. It is found that the ductile-brittle transition temperature increases linearly with the increasing of neutron fluence, and then rises slowly and finally saturates. The effect of neutron flux rate on the embrittlement of RPV materials is not obvious.