【Objective】 Studying the response of evapotranspiration to land use change in Chinese Loess Plateau (CLP) is of great significance to clarify the effects of the Grain for Green Project on regional climate. Many studies have focused on the temporal trend or spatial distribution of reference evapotranspiration based on meteorological or satellite data at the watershed or regional scales. However, these studies cannot reflect the actual situation of the impact of land use change on surface evapotranspiration. Thus, the quantitative evaluation of actual evapotranspiration at the stand scale is lacking. Therefore, the objective of this study is to evaluate the effects of conversion from cultivated farmland into apple orchard on the actual evapotranspiration. 【Method】 A paired experimental study was conducted at the Changwu Tableland, south of the CLP. Both the cultivated farmlands and 20-year-old apple orchards were sampled with a soil auger (0.06 m in diameter) at 0.2 m intervals at each of the five selected sites, with a total of 10 deep soil cores (10 m). The volumetric soil water content was calculated by mass water content and soil bulk density, and the chloride concentrations of soil water were measured by extraction method. Since the local apple orchards were all converted from cultivated farmlands and the soil texture is uniform, the space-for-time method was applied. As precipitation is the only source of water for local crops, the chlorine mass balance method can be used to estimate groundwater recharge. Therefore, the surface evapotranspiration can be estimated by combining soil water mass balance with chloride mass balance. 【Result】 The results show that the averaged soil water content of 4～10 m soil layer in 20-year-old apple orchard was 0.20 m³ m^–3. This was significantly lower than that in farmland (0.28 m³·m^–3) and implies that soil water content were obviously affected by deep root after cultivated farmland converted into apple orchard 20 years ago. The long-term averaged groundwater recharge rate was 57±13.5 mm·a^–1 in cultivated farmlands, which resulted in the actual evapotranspiration of 527±13.5 mm·a^–1 and accounting for 90%±2.3% of the annual precipitation. After cultivated farmland converted into 20-year-old apple orchard, the actual evapotranspiration significantly increased, with an average of 625 mm·a^–1, accounting for 107% of the annual precipitation. Compared to cultivated farmlands, the actual evapotranspiration, in total, increased 1960 mm in 20-year-old apple orchard from 20 years ago, with an annual average of 98 mm. Within the 20-year-old apple orchards, soil water in 4～10 m, and 10～18 m soil layers contribute 24 mm·a^–1(4%), and 41 mm·a^–1(3%) to the annual evapotranspiration, respectively. 【Conclusion】 The small contributions of 4～10 m and 10～18 m soil layers indicated that the deep soil water (below 4 m) has an important role in evapotranspiration of the deep rooted apple tree. However, the main water sources for evapotranspiration still depend on the shallow soil water (0～4 m) that is easily recharged by the latest precipitation. This is the first study that quantitatively evaluated the impacts of cultivated farmland being converted into apple orchard on the evapotranspiration and the contribution of deep soil water to evapotranspiration. This study provides a scientific basis for evaluating the effects of land use change on the regional evapotranspiration and climate on the CLP and other regions with a significant land use/cover change.