Soil moisture has large impacts on agricultural production and regional ecosystems. The Loess Tableland was selected as the study area, and the characteristics of soil moisture in different land use types were measured and analyzed from March to October 2012. The objectives of this study were to characterize the dynamic changes in soil water, determine the soil moisture depletion and replenishment depths, and compare the differences in water consumption under the different land uses. The results showed that: (1) Soil moisture content recorded in the 0-600 cm layer in a 7-year-old apple orchard was the highest, followed by that in a corn field, wheat field, and a 17-year-old apple orchard. Soil moisture contents in different land use types had the same variation trends as precipitation during the observation period, but their responses lagged changes in precipitation. (2) Soil moisture variation declined with increasing soil depth in different land use types. The variation in surface soil water content was greater than that of deep soil water content because the surface layer (0-100 cm) was influenced by multiple processes, including rainfall infiltration, soil evaporation, and plant root water uptake. Soil moisture was less affected by outside conditions as soil depth increased; therefore, soil moisture tended to be relatively stable. In addition, because of complicated functions of plant growth regulation and local phenological characteristics, seasonal variation was also observed in the soil water profile in each land use type. The shape of the average soil moisture profile for the 7-year-old apple orchard was an S-type curve, while the soil moisture profile for the 17-year-old apple orchard showed an initial decrease, followed by an increase and a final decrease. The deep soil moisture stabilized around 13.5%. The shapes of the soil moisture profile in wheat and corn fields were similar; they both increased at first and then decreased. (3) No soil desiccation occurred in farmland or in the 7-year-old apple orchard, while relatively deep desiccated soil layers existed in the 17-year-old apple orchard at depths of 320-600 cm. This was because the water consumption in the 17-year-old apple orchard was greater than water replenishment. In lower rainfall years, the precipitation was not enough to meet the demand from tree growth, so water stored in deep soil was used. Due to the limitation of rainfall infiltration depth, it was hard for water to be replenished once it was depleted. Thus, desiccated soil layers formed under continuous dry years. (4) There were some differences between the soil moisture depletion and replenishment in different land use types. The depths of soil water depletion in the 7-year-old apple orchard, corn field, and wheat field were 200 cm, 300 cm, and 300 cm, respectively, and their replenishment depths were beyond the measured depth of 600 cm. Their deep soil moisture conditions got compensation and improvement after the seasonal precipitation. The depth of soil water depletion in the 17-year-old apple orchard was 500 cm, while the replenishment depth was about 200 cm. As the replenishment depth was less than the consumption depth, the desiccated soil layers already present in the 17-year-old apple orchard will continue to exist for some time if there is no human intervention.