Abstract: Digital elevation model（DEM）can quantitatively describe landforms. While the statistic approach of karst entities based on peck-cluster depression has not been established yet. Based on 1:50,000 digital topographic maps of Dawan，Muxiang，Yanzikou and Zhuzhong in the Wumengshan mountains, this work builds the DEM of the study area and the peak-cluster depression DEMs by spatial analysis. On the basis of 1:50,000 geology surveys, the strata in the survey area are reclassified into seven karst strata association types， volcanic strata group, clastic strata group, limestone strata group, dolomite strata group, marl strata group, limestone-clastic interbed group and dolomite-clastic interbed group. Taking karst strata association as a statistical unit, the parameters of peak-cluster depression of each unit are analyzed. Peak-cluster depression development coefficient (k) is defined to quantify the development degree of karst. The geomorphology of the survey area is divided into 5 landscape types. Of them, Ppeak-cluster depression, k＞8.5, peak depression is extremely developed. Ridged-hill valley, 1.9＜k≤8.5, the cone peak is relatively developed, and the depression is moderately developed. Clustered-hill valley, 0.9＜k≤1.9, the cone peak is weakly developed, and the depression is rare. Karst mountain，0＜k≤0.9, cone peak depression is also rare. Middle mountain，k=0, developed gullies, no karst. The karst geomorphologic types established are consistent with the combined characteristics of karst entities. The classification results reflect the distribution and differentiation characteristics of different karst landscape in a small range.