Abstract: The largescale spatial distribution of stable isotopic composition in soil water is jointly controlled by the precipitation isotope and its natural environment, and is vital information in simulation of isotopes in different types of water in land surface process. Considering the deficiency of existing insitu monitoring in soil water isotope on a large scale, GCM (General Circulation Models) is an effective tool to understand the spatial distribution of soil water isotope. Based on the simulations of several GCMs including LMDZ (free), LMDZ (nudged) and MIROC (free), the spatial and temporal characteristics of δ¹⁸O in soil water in China’s land surface were analyzed, and the relationships between soil water δ¹⁸O and other parameters (surface air temperature and δ¹⁸O in precipitation) were also presented. The results showed that: (1) The oxygen isotopes in soil water were enriched in southeastern China and relatively depleted in northeastern, northwestern China and Tibetan Plateau. The lowest values were found in Tibetan Plateau. The isoscape of soil water δ¹⁸O coincided with that of precipitation δ¹⁸O on a national scale. During 1979-2007, the annual, summer and winter means of δ¹⁸O in soil water presented an arising trend. (2) The soil water δ¹⁸O correlated with surface temperature, and a positive correlation was generally detected at high and mid latitudes; the strong positive correlations were seen at the inland areas far away from oceans. (3) The correlations of isotopes in soil water and precipitation simulated in the three data sources were all positive across China, which suggests the precipitation is usually the ultimate origin of soil water. The correlation coefficients were relatively large in northwestern China, and gradually declined from south to north in northeastern China. The correlation coefficient presented a circular distribution in Tibetan Plateau with weak correlations in the center.