Abstract: Urban land supply is an important part of the foundation for urban expansion. Analyzing the spatial pattern of the price of different types of land is of great significance to urban land rational development, and it has become an important perspective from which researchers examine the process of urban expansion. By using the methods of conventional statistics and geostatistics, taking the selling prices of residential, industrial, and commercial land during 2001-2010 as samples, we analyzed the statistical characteristics and spatial distributions of transaction-based land prices in Nanjing City. The results showed that the spatial distributions of selling price are different for different types of land-use. Commercial land price is highly dependent on the extent of business activity of CBD (central business district), and the high value areas are concentrated near the center of the city—Xinjiekou & Confucius Temple District. The land price gradually reduces toward the periphery, showing a typical single-centered pattern. Residential land price is consistent with the level of environmental quality; high value areas have lower level of aggregation by comparison to commercial land, showing a trend of shift from the center to periphery of the city in a clear pattern of circular layers. Industry land price is determined mainly by outbound traffic conditions, environmental costs and the policy effects; the locations are further away from the city center, except for the industrial activities that have high demand of basic infrastructures and other conditions, such as high-tech industries. The distribution of industry land shows a multi-centered pattern.

Key words: Nanjing, semivariance, spatial distribution, spatial interpolation, urban land prices