Because the traditional fault location method of industrial Ethernet network communication line has the problems of high location false alarm rate, low location accuracy rate and long location time, a Bayesian-based industrial Ethernet network communication line fault location method is proposed. First, the fault signal is measured by discrete Fourier transform algorithm. Then according to the measurement results, the fault features are extracted. Finally, the extracted features are input into Bayesian as training samples, and Bayesian is used to calculate the maximum posterior probability output corresponding to the input data, build a fault localisation model, and obtain the localisation results. The experimental results show that the fault localisation rate of the proposed method is only 0.8%, the localisation accuracy rate is as high as 98.6%, and the localisation time is 10.3 s. The fault localisation effect is good, and the fault localisation time can be effectively shortened.