Abstract: The large cluster Dendrocalamus giganteus forest in Cangyuan County, Yunnan Province was taken as the research object, and the GF-2 image was taken as the main information source. Combined with the investigation of 32 sample plots on the ground, the optimal scale selection of region-wide cluster D. giganteus LAI remote sensing estimation was analyzed by using geostatistical variograms method and partial least-squares regression model verification method. The results show that the optimal scale of the cluster D. giganteus is 6.9 m based on the variograms analysis, and the optimal scale based on the PLSR model is 7 m. So the optimal image observation scale of cluster D. giganteus is 7 m. The results of the 2 methods are close and consistent in the selection of spatial optimal scales and can be used as a reference for the selection of optimal scale for remote sensing estimation of other bamboo parameters.