To improve the accuracy of monitoring land gravelization and the efficiency of fieldwork, and further promote the innovation of monitoring methods, a study on the appropriate size and quantity of quadrats was conducted. This study examines the desert region of Inner Mongolia and uses the entropy TOPSIS and Wilcoxon rank sum test methods to determine the optimum sampling scheme for land gravelization monitoring. Based on results using 100 cm×100 cm quadrats as the true values of gravel coverage and surface gravel mass per unit area, new survey areas in the desert regions of Inner Mongolia were selected, and the maximum quadrat was expanded to 200 cm×200 cm to increase the sampling areas and number of quadrats, and optimize the sampling scheme for monitoring land gravelization. The comparative results of the two experiments show that: \mathrm{①} Q25 exhibits good advantages for monitoring suitability, which can improve field work efficiency while ensuring measurement accuracy; \mathrm{②} the appropriate quantity and sizing of quadrats vary significantly with changes in the maximum quadrat size; however, for Q25, the change in appropriate quantity (from 9 to 10) is non-significant, indicating good stability; \mathrm{③} using Q25, the appropriate coverage for monitoring gravelization is consistent with the results from monitoring surface gravel mass per unit area; this not only simplifies the monitoring process, but also ensures the reliability and comparability of monitoring data.