Rapid acquisition of soil moisture content (SMC) in crop root zone is the key to drought supervision and precision irrigation. The relationship between the unmanned aerial vehicle (UAV) multispectral remote sensing and SMC was mainly studied based on the field maize data of experimental station in Zhaojun Town, Dalate Qi, Inner Mongolia. The canopy images of field maize with five irrigation treatments were obtained at different growth stages (vegetative stage, reproductive stage and maturation stage) through the six-rotor UAV equipped with 5-band multispectral camera, and the SMC values at corresponding time was acquired by drying method on the field at five soil depth (10cm, 20cm, 30cm, 45cm and 60cm). Then the spectral reflectance of field maize canopy was extracted to calculate a number of vegetation indexes (VIs). Firstly, data was adopted to analyze the grey relationships between SMC and the selected typical VIs，and the selected typical VIs were used to determine the sensitivity of different VIs to SMC at different growth stages. Secondly, machine learning models of Cubist, back propagation neural network (BPNN) and support vector machine regression (SVR) were constructed and verified. The result showed that the three machine learning models showed good performance on modeling and prediction at different growth stages. The effectiveness of the SVR model was optimal among the three machine methods. The effect of the BPNN model followed, and the Cubist model was relatively the worst. The optimal model was the SVR model at M stage, the modeling R2 and validation R2 for the SVR model were 0.851 and 0.875, and the root mean square error (RMSE) both were 0.7%, and the normalized root mean square error (nRMSE) were 8.17% and 8.32%, respectively. The inversion accuracy of the SVR model at R stage performed badly, the modeling R2 and validation R2 for the SVR model were 0.619 and 0.517, respectively. The research result was of great significance to monitor the soil moisture content in root area of crops and meaningful to precision irrigation.