Abstract: Taken Shuangjiang county as an example, the impacts of Jatropha curcas energy forest on landscape pattern in different altitudes were analyzed. The results showed that:(1)Dominant landscape types were farmland, warm temperate coniferous forest and semi humid evergreen broad leaved forest, including 15 types of vegetation landscapes and 3 types of non-vegetation landscapes in the study area.(2)The dominant landscape types were different at each altitude. At altitude above 2 500 m, there was mainly mountainous humid evergreen broad-leaved forest.(3)At landscape construction level, the landscape pattern displayed a regular change along with change of altitudes. With the altitude increasing,the change curves of total area(TA), number of patches(NP), landscape shape index(LSI), area weighted average shape index(SHAPE_AM), area weighted average fractal dimension(FRAC_AM), landscape richness(PR)were “∩”shape, but landscape richness density index(PRD)was a “∪”shape change. And patch density(PD), Shannon diversity index(SHDI)and Shannon evenness index(SHEI)decreased gradually with elevation increase. But contagion index(CONTAG)and aggregation index(AI)increased.(4)Because of the afforestation, dominant position of hot shrub landscape raised, and it become more concentrated, and the shape of it was more regular. The landscape types located in the higher altitude and the non vegetation landscape were not influenced by afforestation. But for the other landscape types, the dominance position reduced, and the patch was more fragmented.(5)J. curcas energy forest only affected landscape pattern at the altitude below 2 000 m, but didn’t cause the change of the landscape types, which made the landscape been more fragment, and landscape shape been more regular. Landscape at the altitude below 1 000 m was significantly affected, at where hot shrub landscape became dominant,and landscape at altitude of 1 000-1 500 m was less influenced than that at altitude below 1 000 m,landscape at altitude of 1 500-2 000 m was least affected.