This paper presents a Probabilistic Neural Network (PNN) for loom system fault diagnosis. Using the known data to train the probabilistic neural network. Then, input the samples of loom system to diagnose the fault type. In this paper, we display some common fault samples according to related datas. In order to prove the accuracy of the PNN for fault diagnosis of loom system, we use MATLAB to develop PNN program. The simulation can prove the proposed method is effective and accurate.