The invention provides a low-voltage transformer area operation state comprehensive evaluation method based on a neural network model. The method comprises the following steps: S1, obtaining a user average active power value of a plurality of historical transformer areas at an interval of one hour in one day; S2, classifying the transformer areas by utilizing an improved K-Means clustering algorithm according to the average active power of the users in the transformer areas; S3, analyzing and calculating evaluation index weight values influencing the operation state of the transformer area byutilizing an order relation; S4, calculating according to the weighted value to obtain a historical transformer area operation state evaluation value; S5, performing model training by taking the historical transformer area operation state evaluation value as an RBF neural network sample to obtain an RBF neural network low-voltage transformer area operation state evaluation model; and S6, utilizingthe trained RBF neural network low-voltage transformer area operation state evaluation model to carry out operation state evaluation on different types of transformer areas. By means of the method, comprehensive and accurate evaluation of the operation state of the low-voltage transformer area can be achieved, problems existing in the transformer area are found, field operation and maintenance personnel are guided to carry out troubleshooting, and the lean management level of the transformer area is improved.