Abstract: Based on a grid (20 m × 20 m) sampling, spatial heterogeneity and pattern of soil nutrients in sloping field in the gorge karst region, southwestern China, were explored by using classical statistics and geostatistics methods. The results showed that soil nutrient contents in slope field in the canyon karst region were more abundant, where pH value had a weak variation and the soil organic matter (SOM) had a moderate degree of variation. All the soil nutrients had moderate or strong variation with an order of available phosphorus (AP) > total potassium (TK) > SOM > alkaline nitrogen (AN) > total nitrogen (TN) > total phosphorus (TP) > available potassium (AK). All of the soil nutrients had a good spatial autocorrelation and the autocorrelation function performed in the same law of developing from positive to negative direction with the inflection point ranged from 80 to 100 m. In addition, the Moran’s I was small for TK and AP while large for other nutrients. Characteristics of spatial variation differed among soil nutrients. Exponential model fitted best for TK and AP, in which the ratio of nugget to sill (C₀/(C₀+C)) and the range (A) were small and the fractal dimension (D) was high, showed a strong spatial correlation. Spherical model fitted best for other soil nutrients, with C₀/(C₀+C), the range (A) and D showing a moderate autocorrelation. Kriging analysis clearly indicated that pH, SOM, TN, TP and AN were distributed in a concave pattern, while AP and AK had fragmented patch distribution. Therefore, vegetation, topography, human disturbance and strong heterogeneity of microhabitats are main factors leading to the differences in patterns of soil nutrients on the sloping land in the gorge karst region.