At present most of the subdivision algorithm are 1-4 subdivisions. as the number of subdivision increases, the grid too-rapid growth in the number of patch and die after the breakdown of huge volume of data make it difficult to deal with, we proposed a radical Doo-Sabin Mode adaptive subdivision algorithm addressing the problem. As a measurement criteria, the average vector of the vertex and the angle between the intersecting surfaces of the vertex are used to divide the surface, and then make local subdivision. In this way, when the times of subdivision are fewer (the demand of smoothness is not too high) , the effect of subdivision has little difference , but efficiency of the algorithm can be greatly improved.