Abstract: To explore the effect of interpolation parameters on Disjunctive Kriging (DK) interpolation on DEM precision, the authors took a typical gully of dry-hot valley in Yuanmou, Yunnan as an example, based on the field measured point cloud data, DEM was built by changing kernel functions, search points, search direction and smoothing factors of DK interpolation, and the DEM interpolation accuracy was evaluated by cross validation method and the verification of test data sets. The results showed that: (1) the DEM interpolated with Stable function was optimal in comparison with other functions; (2) the optimal search points was 2 to 5; (3) in comparison of different search directions, the DEM interpolated with directional search had higher interpolation accuracy than that of the DEM interpolated with non-directional search, and interpolated with four-directions was better than that of eight-directions; (4) the DEM interpolation error increased obviously when smoothing factor ≥0.8. By the optimization of interpolation parameters in DK interpolation method, the DEM with the high precision of 0.08 m was produced, which could provide references for the studies on gully.