Abstract: To achieve rapid and intelligent prediction of coal quality,a radial basis function （RBF） neural network prediction model based on fuzzy clustering was established. Based on measured contents of total moisture of as received basis,ash of as received basis,volatile matter of as received basis and total sulfur of as received basis in coal,fuzzy clustering was made; according to the results,several quantitative analytical models RBF based on neural network were established. The model can predict volatile matter of dry ash- free basis,total sulfur of air dried basis,low calorific value of as received basis and gross calorific value of dry basis in coal. And the comparison between the model and the direct use of RBF neural network model was made. The experimental results showed that the analytical model has high precision,strong generalization ability and good robustness.