The invention discloses a composite insulator hydrophobicity grade determination method based on a probabilistic neural network. The improved shape factor method is used as a common composite insulator hydrophobicity characteristic parameter extraction method, the shape factor of the maximum water drop (or the water film) is combined with the area of the maximum water drop (or the water film), sothat the insulator grade judgment result is greatly improved, and the insulator grade judgment result is totally composed of four layers of structures, namely an input layer, a mode layer, a summationlayer and an output layer. The learning process is relatively simple, and the convergence problem is not involved. The input layer and the summation layer can be determined according to actual conditions before learning, and the structure of the probabilistic neural network can be determined only by determining the number of learning samples.