Geostatistics combined with Geographical Information System (GIS) was applied to reveal the spatial varibility of organic carbon inclucling total organic carbon, particulate carbon, readily oxidizable organic carbon and black carbon in the top soil (0~1Ocm) in Shenyang City, and the spatial distribution maps were drawn up. The results showed that the indicators of soil organic carbon were in line with skewed distribution, after the logarithm transform conformed to the normal distribution. The theoretical semivariogram models of total organic carbon, particulate carbon,readily oxidizable organic carbon and black carbon were best described with a liner model, and it was found that spatial variabilities of all soil organic carbon indicators were significant (The ratios of nugget to sill of those availables were 98.1%,100%,97.1%,99.0%). And the spatial autocorrelation distances for them were 16.861km, 21.805 km,15.010km and 18.735km, respectively. Particulate carbon displayed the pure block gold effect and non-spatial correlation. The spatial variability of soil organic carbon mainly was caused by the randomness (human activities, etc).The results could provide the basis for the reasonable use of urban soilt.