Schema matching is a basic problem in many database application domains, such as data integration, E-business, data warehousing and semantic query processing. Recently it has become a research hotspot, and most of the achievements are about the use of element’s own information. Research about element’s own information is very mature at present. As an important piece of information in a schema, structure information can be useful information for schema matching, but the research of structure information is far behind that of element’s own information. This paper divides the similarity between two elements into linguistic similarity and structural similarity, and gets the structural similarity by a new statistic method, and then gets the matching probability by integrating the linguistic similarity and structural similarity. At last, the paper gets the mapping between schema elements according to the matching probability. Extensive simulation experiments are conducted and the results show that this algorithm is better than other algorithms in various performance metrics.