Abstract: On the basis of the spatial characteristics of GIS, an alternative viewpoint for defining and discovering spatial outliers from GIS is proposed,in which the spatial location of a spatial outlier is significantly far from other spatial objects in its neighborhood determined by non-spatial attribute.Then the SOD algorithm is proposed and analyzed in detail.Experimental results on synthetic datasets demonstrate that the proposed approach can effectively and efficiently identify spatial outliers in large spatial data sets.Finally,the authors give the general procedure to mine the spatial outliers from GIS database.