Efficient agricultural water use is crucial for food safety and water conservation on a global scale. To quantitatively investigate the agricultural water-use efficiency in regions exhibiting the complex agricultural structure, this study developed an indicator named water footprint of crop values (WFV) that is based on the water footprint of crop production. Defined as the water volume used to produce a unit price of crop (m3/CNY), the new indicator makes it feasible to directly compare the water footprint of different crops from an economic perspective, so as to comprehensively evaluate the water-use efficiency under the complex planting structure. On the basis of WFV, the study further proposed an indicator of structural water-use coefficient (SWUC), which is represented by the ratio of water-use efficiency for a given planting structure to the water efficiency for a reference crop and can quantitatively describe the impact of planting structure on agricultural water efficiency. Then, a case study was implemented in Xinjiang Uygur Autonomous Region of China. The temporal and spatial variations of WFV were assessed for the planting industries in 14 prefectures and cities of Xinjiang between 1991 and 2015. In addition, contribution rate analysis of WFV for different prefectures and cities was conducted to evaluate the variations of WFV caused by different influencing factors: agricultural input, climatic factors, and planting structure. Results from these analyses indicated first that the average WFV of planting industries in Xinjiang significantly decreased from 0.293 m3/CNY in 1991 to 0.153 m3/CNY in 2015, corresponding to an average annual change rate of –3.532％. WFV in 13 prefectures and cities (with the exception of Karamay) has declined significantly during the period of 1991–2015, indicating that agricultural water-use efficient has effectively improved. Second, the average SWUC in Xinjiang decreased from 1.17 to 1.08 m3/CNY in the 1990s, and then declined to 1.00 m3/CNY in 2011–2015. The value of SWUC was highly consistent with the relative value of WFV in most prefectures and cities, showing that planting structure is one of the primary factors affecting regional agricultural water-use efficiency. Third, the contribution rate of WFV variations from human factors including agricultural input and planting structure was much more significant than that from climatic factors. However, the distribution of agricultural input and the adjustment of planting structure significantly differed among prefectures and cities, suggesting regional imbalances of agricultural development. This study indicated the feasibility and effectiveness of controlling agricultural water use through increasing technical input and rational selection of crops in the face of impending climate change. Specifically, we concluded that, the rational application of chemical fertilizers, the development of the fruit industry, and the strict restriction of the cotton industry should be implemented to improve the agricultural water-use efficiency in Xinjiang.