Highly changeable coal quality, low combustion efficiency, and huge consumption of coal are the three common problems facing most of China's coal-fired power plants. Proper prediction for the coal burnout in power boilers has important economic and environmental significance. By combining critical factors in proximate analysis, coal petrography, and the size distribution of pulverized coal, this article aims to propose a comprehensive method to predict the burnout of pulverized coal in power boilers. Results show that: (a) volatile content and calorific value are the two most important factors in proximate analysis, (b) the most important factors in coal petrography are the relative content of reactive and unreactive macerals and the mean and standard deviation of vitrinite reflectance, and (c) the 90-200-mu m-size fraction is a particularly important factor in the size distribution of pulverized coal. Compared with any parameters or indices based on one of the three aspects in proximate analysis, coal petrography, and pulverized-coal size distribution, the synthetic index into which these critical factors are combined has a better significant correlation with coal burnout.