Abstract: This paper proposes a new method of black position identification of urban road traffic accident by using spatial clustering analysis model based on point and plane models .The proposed method is based on the kernel density es-timation ,count data model ,and GIS technology .The proposed method can mine the spatial data and visualize the analysis results .Results from the crash data show that there are two spatial clusters of high values .The first one has the unit standardization Z value above 2 .58 ,which corresponds to the first level of accident black points .The second has the unit standardization Z value between 1 .96 and 2 .58 ,corresponding to the second level of accident black points .