In this paper five main reality problems which appear in the research on landslides nowadays are discussed, and possible ways and methods to solve them are suggested as follows.

A. Distinguishing landslides. New, old and ancient landslides can be distiagusished by analysing the formation condition of landslides and the deformation signs of slopes. The criteria for determining landslides are listed in Table 1,2.

B. Determining the main trigger factors. The factor change analysis method can be used in solve this problem. Besides, the author emphasizes the importance of the erosion of river and gully.

C. Determinging and linking the slip suriace. Many ways and methods should be used to determine the slip surface, such as analysing structure conditions, determining the direct crop of slip surface, indirectly determining the depth of slip surface, determining the dip angle of slip surface of the main sliding body, comprehensively anslysing the prospecting information, analysing the surface displacement observation data, and observing bore holes etc. Then, the longitudinal and transversal sections of slip surface can be linked according to the characteristics of landslides themselve.