Field data were sampled by the method of contiguous grid quadrate. The spatial patterns of Pseudolarix amabilis populations were studied by the following methods: DispItalic textersal index (C), David moores index (I), Lloyd index (m*/m), Morisita index (Iδ), Parameter of negative index (K) and Cassie index (1/K). The quadrate variance analysis, Greig-Smith and Kershaw methods were used to study the spatial distribution pattern, pattern scale and pattern intensity of Pinus armandi population at different sizes. The result showed that P. amabilis population is most clumped. The result indicated the pattern intensity was decreasing with the size increasing and a single peak value appeared in the curve with the area increasing. There was a slow change in the pattern intensity of population, which means a smaller variation in environment and a bigger difference in the relative density of population patches during diffusion process.