[Objective] The relationship between soil moisture content and the apparent thermal inertia was examined based on the data collected from Wushen Banner of Inner Mongolia, to improve the monitoring accuracy of soil moisture content and to make the results more convincing and reliable. [Methods] Multi-band MODIS remote sensing data and ATI(apparent thermal inertia) method were used. A point-repeated ground sampling scheme was adopted to diminish the poor representativeness of single sampling method in a point. [Results] It was indicated that the correlation coefficient in the present design was significantly higher than that of single sampling method. For the relationships observed in the soil layers of 0-10 cm, 0-20 cm and 0-30 cm, the coefficients were all high, with values of 0.587, 0.658 and 0.650, respectively. The above regression models had different goodness with their averaged relative errors in predicting the measured water content. The averaged relative errors were 21.53%, 26.67% and 22.83%, respectively. [Conclusion] Under point-repeated sampling scheme, the ATI-based monitoring results of water content in Uushen Banner were more reliable.