<p>Weak planes commonly exist inside rock masses and may lead to anisotropic behavior, which influences the shape and volume of the sliding body of a landslide. Moreover, the landslide volume significantly influences landslide runout behavior. Landslide runout prediction is essential for disaster prevention and management. A case study of the Hsiaolin landslide, in which a village was buried by more than 20&#xa0;million m3 of landslide debris in 2009, is analyzed using numerical methods. The anisotropic model is used in finite element analysis to estimate the sliding body. The estimated sliding surface is then incorporated into the runout simulations using the continuum method and the discrete element method (DEM). In particular, the DEM model is calibrated using a large-scale numerical simple shear test to reflect the anisotropic strength of the slope mass from a macroscopic perspective. The anisotropic model can better estimate the landslide volume, and both methods can simulate similar maximum deposition heights, landslide dam heights, and maximum travel velocities. However, the continuum method simulates landslide runout more like a fluid, whereas the DEM simulates landslide runout more like a cohesive solid in the first half of the travel duration. Compared with the DEM, the continuum method simulates that the landslide debris reaches the downslope river earlier but takes longer to cease movement. Furthermore, the simulated collapsed geomaterials spread over longer distances and show greater variation in flow height in the first half of the travel duration. Satisfactory simulations are performed to match the actual deposition area and travel time. For prediction purposes, different scenarios associated with the assumption of the model parameters should be considered owing to the uncertainty of the parameters in the two methods.</p>