To achieve rapid identification of Ganoderma lucidum medicinal materials from different places of production and cultivation methods, a high-sensitivity X-ray fluorescence spectrometer is used to determine 52 Ganoderma lucidum medicinal materials. 52 samples can be divided into 5 categories through systematic clustering method. Based on this, a Gaussian mixture model is used to verify and analyze the clustering categories, and a multi-layer perceptron is used to train the classification model. Results show that the detection accuracies of both training and test sets by the multi-layer perceptron are 100%. The combination of high sensitivity X-ray fluorescence spectroscopy and chemometric methods can quickly and effectively classify and test Ganoderma lucidum medicinal materials, providing a technical method for rapid testing of food, drugs, and environment cases using Ganoderma lucidum as evidence.