Through evaluating a few existing multi-source products of gross primary productivity (GPP) and evapotranspiration (ET), we constructed combined datasets of GPP and ET for the Qinling Mountains in China. Then, the spatial-temporal variations of water use efficiency (WUE) were calculated and analyzed for the Qinling Mountains. The results indicated that the estimation effects of different products of GPP and ET were very different for different vegetation types. Generally, the products of VPM GPP, RF GPP performed well for GPP estimation, while the products of China ET and PML-V2 ET performed well for ET estimation in China. In the Qinling Mountains, the average annual ecosystem GPP_year was about 1 489.95 g C/m²; the average ET_year was about 588.49 mm; the average WUE_year was about 2.56 g C/kg H₂O; and the change rate of average WUE_year was about 0.156 g C/(kg H₂O·10 years). The values of average WUE_year showed a non-significant rising trend (p > 0.05). The differences of annual average WUE_year were not great among different vegetation types. However, the trends of WUE_year changes were various for different vegetation types. Some vegetation types showed increasing trends, while others showed decreasing trends. The dynamics of average montly WUE_month had two obvious peaks per year. This study provided a reference for data acquirement and selection for the research of carbon-water relationships in different ecosystems. It also deepened our understanding of the temporal and spatial changes of ecosystem WUE in the Qinling Mountains in China.