Choosing the optimal spatial interpolation method is of great significance to clarify the spatial distribution of soil organic matter in arid areas under different degrees of disturbance. When sampling soil in Fukang City, Xinjiang, the Ordinary Kriging (OK) method, the inverse distance weight (IDW) method, the radial basis function (RBF) method, and the local polynomial (LPI) method were used to describe the spatial distribution characteristics of soil organic matter in arid areas under different degrees of human disturbance, so as to explore its optimal spatial interpolation method. Our analyses showed that:(1) When the degree of human disturbance was on soil was great and the content of soil organic matter was low, the spatial variability ranged from weak variation to medium variation, and the coefficient of variation increased from 9.27% to 28.7%. Meanwhile, the spatial autocorrelation of soil organic matter decreased gradually, showing strong, moderate and weak correlation from an unmanned disturbance area, a human disturbance area, and to a severe human disturbance area, respectively. The stronger the degree of human disturbance, the greater the effect of random factors on soil organic matter. (2) The interpolation accuracy of the four methods decreased with the increase of the interference degree. However, the OK method had the best interpolation effect on the unmanned disturbance area with strong spatial structure; the R² was 0.625. The interpolation accuracy of the RBF method was the highest in the human disturbance area and the severe disturbance area; the R² were 0.562 and 0.434, respectively. These results provide a scientific reference for finding a suitable spatial interpolation method for soil organic matter in arid areas with different degrees of disturbance.