Abstract: Spatial distribution of soil medium- and micro-element in the middle plain of Jilin Province were investigated using geostatistics and geographic information system (GIS)techniques. With various processing methods, including logarithmic transformation, outlier exclusion and Box-Cox transformation, available Ca, Mg, B, Cu, and Zn contents were evaluated with normal distribution tests.The result shows that Box-Cox transformation is effective in achieving normal distribution data sets for available Ca, Mg, Cu, and Zn, while outlier exclusion is effective for available B weakening effect by outliers. Geostatistical analysis including calculation of experimental variograms and model fitting were conducted, and the distribution patterns of the soil nutrient elements plotted. Semivariogram analysis shows that the soil nutrient elements are moderately spatially correlative within a given range.