Abstract: In order to explore the distribution pattern of the main soil mineral components in karst cluster-peak depression area and to guide vegetation restoration and ecological reconstruction in the area, the spatial heterogeneity of mineral components (SiO2, Fe2O3, CaO, MgO, Al2O3, and MnO) in surface soil (0-20cm) in Mulun National Nature Reserve was studied by the methods of classical statistics and geostatistics. The soil samples were collected with grid method based on a dynamic monitoring plot (200 m × 100 m) in Mulun National Nature Reserve karst cluster-peak depression area. The results showed that the content differences of the six mineral components were large, and the sum of SiO2 and Al2O3 accounted for 89.53% of the total six mineral components. The variance coefficients of all the six mineral compositions were large. The spatial patterns of the six mineral components were quite different from each other, and fit to different models of mineral components. Both SiO2 and MnO had medium spatial autocorrelation with long range and well spatial continuum. Fe2O3, Al2O3, CaO, and MgO were characterized by strong spatial autocorrelation with short ranges. The Kriging contour maps indicated that the distribution pattern of CaO and MgO were similar, but opposite to that of SiO2. And the distribution pattern of Fe2O3 was similar with that of Al2O3, while MnO was in a relative homogeneous distribution status. The results indicated that topography, micro-physiognomy, precipitation, the history of human disturbance, and especially vegetation were the most important factors that affecting the spatial patterns of soil mineral components in karst cluster-peak depression region. Therefore, increasing plant diversity and vegetation coverage would be of great importance for effective improvement and rational utilization of soil mineral resources.