Abstract: The pore structure measurement in coal is commonly analyzed by mercury intrusion method, low temperature liquid nitrogen method and nuclear magnetic resonance method. Due to the different test methods, the results cannot be used uniformly, and the matrix compression effect in mercury intrusion method will cause large errors. In order to solve this problem, three high-rank coal samples from Jincheng and Changzhi Mining Areas in Qinshui Basin were taken as examples, and used mercury injection, low-temperature liquid nitrogen adsorption and nuclear magnetic resonance relaxation methods to test the pore structure of coal samples respectively. Through the compressibility correction of mercury injection data and splicing with low-temperature liquid nitrogen data at the connection aperture, the pore structure of coal was characterized by joint characterization, and the pore size distribution characteristics of coal were compared and analyzed in combination with nuclear magnetic resonance. The results show that the pore volume corrected by mercury intrusion data is closer to that of low temperature liquid nitrogen, and the deviation value is 22.40%-38.51%; the pore structure of different coal samples is analyzed by the joint pore method, and the position of the joint pore is between 75-89 nm; the results show that the pore volume of coal samples is 0.001 36-0.004 58 cm³/g, and the pore volume ratio of different pore sizes is transition pore>macropore>mesopore>micropore; compared with the single test method, the pore size distribution of the combined pore method and the nuclear magnetic resonance method is closer, but there are also differences. The average distribution ratios of micropores, transition pores, mesopores and macropores characterized by the combined pore method are 8.68%, 45.58%, 20.54% and 25.20%, respectively. The average distribution ratios of micropores, transition pores, mesopores and macropores characterized by the nuclear magnetic resonance method are 10.64%, 64.21%, 14.23% and 10.92%, respectively. The reason for the difference may be mainly related to the pore structure of coal changed by mercury intrusion method and low temperature liquid nitrogen method; through the comparison of NMR results, it is found that the combined corrected mercury injection and low temperature liquid nitrogen data can improve the accuracy of the pore size distribution test results in coal.