Rapid and non-destructive monitoring of soil moisture content in farmland are essential contents of accurate irrigation. Traditional soil water measurement methods are applicable to small scale and difficult to obtain spatial continuous soil moisture information. Soil moisture is affected by microtopography and soil texture. It has strong spatial variability characteristics. Electromagnetic induction technology has been widely used in soil properties investigation owing to its advantages of rapid, efficient, and non-destructive. In this study, the cotton fields under mulched drip irrigation in Alar National Agricultural Science and Technology Park in southern Xinjiang are considered. EM38-MK2 is used to obtain soil apparent electrical conductivity data of four groups of cotton fields at different stages rapidly and effectively. Surface soil samples (0-20 cm) are collected simultaneously. The observation points’ water content is obtained by constructing an inversion model between the soil apparent electrical conductivity data and the indoor measured water content data. Then, the study area’s soil moisture is divided into the soil moisture and drought classification criteria. Finally, the soil moisture’s spatiotemporal variability is investigated using GIS software and geostatistical methods synthetically. The results show that the determination coefficients of the soil moisture inversion model in four different periods are greater than 0.80, and root mean square error (RMSE) and mean absolute percentage error (MAPE) are small, indicating that the inversion model has high precision and the correlation between soil apparent electrical conductivity and surface soil moisture is good. The data of soil moisture content in different periods show that moisture has strong time variability, which changes from medium variability to weak variability and then to medium variability. Moreover, the variation function model is different owing to the influence of artificial irrigation factors. In the semi-variance analysis, the ratios of Nugget and Sill of soil moisture in four different periods are more than 75%, indicating that soil moisture tends to be a weak spatial correlation. The maps of elevation inverse distance weight (IDW) interpolation and moisture Kriging interpolation show that microtopography is an essential factor affecting soil moisture distribution. This study can provide an important method of support for the dynamic monitoring of soil moisture in cotton fields under mulched drip irrigation in arid areas to better guide agricultural irrigation.