The secondary forest of Mt. Longhu in the south Dongting Lake is chosen as the research object in this paper which tries to establish a stand model based on Voronoi diagram. This diagarm is applied here to illustrates the stand spatial structure and extract forest spatial quantitative information. Based on this, coefficient of variation (CV) is used to quantify the change rate of Voronoi polygon area, transferring the analytic results of stand spatial pattern to the calculation of the CV based on Voronoi diagram, and comparing the calculation results with uniform angle index. The result shows that the stand model of Voronoi diagram can accurately display some important spatial information, such as the correlation, distance and angle between adjacent trees as well as spatial density, etc. When sampling large scale(above 76m) , CV tends to remain stable, and the overall stand spatial pattern shows mild aggregation which is close to random distribution. The CV of Voronoi diagram and uniform angle index have notable positive correlation, which further proves the feasibility and effectiveness of the analytical methods applied in this paper.The spatial pattern of main tree species of abor layer in different development stages shows a distinct temporal and spatial dynamic characteristics due to many factors. The small鄄diameter wood belongs to aggregated distribution, while the large鄄diameter belongs to mild aggregation or random distribution. The stand spatial model and its pattern analytical method, both based on Voronoi diagram, charaterize the tree growth feature, intraspecific and interspecific relationship, characteristics of resources and enviornment distribution from the perspective of spatial structure, providing a new way of thinking for stand spatial management by applying computational geometry method.