Here, we have synthetized 16 tree-ring delta~(18)O series from western part of China; most of sampling sites are distributed on the Tibetan Plateau and the arid region of Northwest China. These sites are mainly near to the local forest upper treeline. Tree samples we selected are from the regional dominant species, including spruce, fir, larch and juniper. We firstly compared the spatial patterns of the mean values of tree-ring delta~(18)O during the period from 1901 to 2004 with the mean values of delta~(18)O in precipitation, and find that tree-ring delta~(18)O share the similar spatial distribution pattern with precipitation delta~(18)O. In addition, tree-ring delta~(18)O shows relative lower values in the southern Tibetan Plateau, while a higher value is revealed in the northern Tibetan Plateau. Tree-ring delta~(18)O values increase along with the latitude of the tree-ring delta~(18)O sampling site, which is similar to the latitude effect of delta~(18)O in precipitation. The climate responses between tree-ring delta~(18)O and meteorological dataset from the nearest tree-ring sampling sites reveal that delta~(18)O in tree rings is significantly positively correlated with the growing season temperature, and is significantly negatively correlated with the growing season precipitation and relative humidity. By rotated empirical orthogonal function analysis and spatial response analysis,we point out that in northern part of Xinjiang, where climate is mainly dominated by the westerly circulations, tree-ring delta~(18)O reflects mainly the regional drought (moisture) conditions; in the southeastern Tibetan Plateau, where climate is dominated by the summer monsoon, tree-ring delta~(18)O records the regional hydroclimate variability ; while in the intersection area between monsoonal area and westerly area, tree-ring delta~(18)O mirrors climate signals relative weakly. Our results suggest that tree-ring delta~(18)O can be an excellent proxy in regional climatic reconstruction based on it's better ability in keeping the long-frequency climatic signals over larger region. The reinforcement of the ongoing tree-ring isotopic network over the entire China will be helpful to understanding the regional climate change which is linked by differing atmospheric circulations systems and forest dynamics under the significant climate changes in the future.