Utilising datasets from the Global Network of Isotopes in Precipitation of the International Atomic Energy Agency, and previous isotopic studies, we investigated δ18O spatial and temporal patterns in Chinese precipitation. Significantly positive relationships existed between precipitation δ18O and air temperature above the north of 35°N and in high altitude regions above 32°N. Significantly negative relationships between precipitation δ18O and the precipitation amount existed below south of 35°N. These temperature and precipitation effects became stronger with increasing altitude except in high altitude regions between 32°N and 35°N. The NCEP/NCAR reanalysis data from 1980 to 2004 showed that variations in spatial and seasonal wind fields at 700 hpa and total precipitable water from the ground to the top of the atmosphere were correlated with the monthly spatial distribution of precipitation δ18O. Basing on this relationship, we established quantitative correlations between the mean monthly precipitation δ18O and both latitude and temperature in different seasons. We found that spatial variations in precipitation δ18O could be described well using the Bowen–Wilkinson model and second-order equations developed during the present study only in winter (from December to February). During the rest of the year, patterns were too complex to predict using simple models. The results suggest that it is difficult to demonstrate variations of precipitation δ18O throughout the year and for all regions of China using a single model. Moreover, the new models for the relationships among precipitation, latitude, and temperature were better able to depict the variations in precipitation δ18O than the Bowen–Wilkinson model. Copyright © 2011 John Wiley & Sons, Ltd.