Landscape pattern is an important measure of aesthetics and its environment. Understanding its changes spatially and temporally and their driving forces is of great intent for environmental change studies. In this study, land use/cover data in Xing long mountains area in 1995 and 2000 were analyzed to quantity the landscape changes and to identity the driving forces. Specifically, RS, GIS and FRAGSTATS software were used to compute the landscape pattern indexes for 1995 and 2000; SPSS software was used to the relation among the landscape indexes and TAPE-G, SRAD and WET methods were used to analyze the natural driving force of the spatial distribution of SHDI. The results showed that meadow and crop-land played a dominant role in the study area. The meadow and crop-land had an increasing trend, while forest and water-body decreasing from 1995 to 2000. More over, SHDI was increasing from 2.1156 to 2.1242, LPI was decreasing from 5.145 to 5.1038, AI was increasing from 95.6533 to 95.6639 and DIVISION was decreasing from 99.385 to 99.3717 between the two periods. Secondly, the correlation coefficient between the SHDI and SHAPE-MN was very low, while those of the other landscape indexes were high in the 0.01 level. Thirdly SOILWATER and STEADYWET, the outputs form the WET program, evidently influenced the spatial distribution of the landscape index SHDI, when no human activity was considered.