Abstract. An accurate assessment of soil water budget components (SWBCs) is necessary for improving irrigation strategies and optimizing the use of fertilizer in any water-limited environment such as the desert oases in arid northwestern China. However, quantitative information of SWBCs is usually challenging to obtain, because, since the water cycle is principally driven by irrigation (I), drainage (D), and evapotranspiration (ET) in desert oasis settings, none of the drivers can be easily measured under actual conditions. Soil moisture is a variable that integrates the water balance components of land surface hydrology, and the evolution of soil moisture is assumed to contain the memory of antecedent hydrologic fluxes, and thus can be used to determine SWBCs from a hydrologic balance. A database of soil moisture measurements from six experimental plots in the middle Heihe River Basin of China (NT1 to NT6, designed to investigate the long-term effects of cropping systems and agronomic manipulation on soil property evolution in the ecotone of desert and oasis) was used to test the potential of a soil moisture database in estimating the SWBCs. The experimental plots were treated as continuous pasture cropping, maize cropping, maize cropping with straw return, maize-maize-pasture rotation, maize-pasture rotation, and maize-pasture intercropping. We first compared the hydrophysical properties of the soils in the plots, including soil bulk density (ρ_b), vertical saturated hydraulic conductivity (K_s), and soil water retention features, and then determined evapotranspiration and other SWBCs through a data-driven method that combined both the soil water balance method and the inverse Richards function. Our results showed that although the tillage and planting of the past decade have significantly increased the soils' water-holding ability, the magnitude of increase in most of the parameters was independent of the treatments applied across the plots. Despite the relatively flat topography and consciously uniform irrigation, significant variances were observed among the plots in both the cumulative irrigation volumes (between 652.1 mm at NT3 and 1186.5 mm at NT1) and deep drainages (between 170.7 mm at NT3 and 651.8 mm at NT1) during the growing season of 2016. Obvious correlation existed between the volume of irrigation and that of drained water. However, the ET demands for all the plots behaved pretty much the same, with the cumulative ET values ranging between 489.1 and 561.9 mm for the different treatments in 2016, suggesting that the irrigation amounts had limited influence on the accumulated ET throughout the growing season. This work also confirmed that relatively reasonable estimations of the SWBCs in a desert oasis environment can be derived by using soil moisture measurements, and the results will provide a great potential for identifying appropriate irrigation amounts and frequencies, and thus move toward sustainable water resources management, even under traditional surface irrigation conditions.