Abstract: Spatial clustering under spatial and attribute constraints is the clustering analysis on the spatial dataset with non-spatial attributes,which is named dual clustering.The result of dual clustering should be spatially continuous and attributively aggregative.The essence of dual clustering is to find out the clustering and distribution rules of non-spatial attributes.This paper presents the formalized definition of dual clustering,proposes the two-step strategy and self-organizing dual clustering algorithm.Case study verifies the algorithm and shows that the self-organizing dual spatial clustering can find the spatial distribution rules of non-spatial attributes,such as clustering and stretching.Self-organizing dual clustering can detect clusters with complicated shape and reduces the artificial influence.