Abstract: Aiming at the uncertainty of coal powder fineness with small sample and unknown distribution type, an uncertainty evaluation method based on grey model was proposed. Through the maximum dispersion of pulverized coal fineness measured, the grey standard deviation of pulverized coal fineness was defined, and the grey uncertainty was calculated. The standard uncertainty evaluation methods proposed in this paper and grey uncertainty were used to evaluate the pulverized coal fineness of the same group of coal-fired boilers. The results show that there is no significant difference between the two evaluation methods, which are 0.087 and 0.090 respectively, and the uncertainty evaluation method of grey model is more simple and easy. In the actual working conditions, other conditions of coal-fired boiler are known, the grey model for uncertainty evaluation can easily and accurately obtain the best pulverized coal fineness, which can provide effective operation guidance for operators to timely adjust the combustion process.