The study of coupled relation between grasshopper and vegetation could help revealing the habitat selection and adaptability of the grasshopper and collaborative relationship within populations. The experiment was conducted in the Baidaban grassland which is located in upper reaches of Heihe of the north Qilian Mountain (38°48'0″-38°49' 50″N, 99°37'15″- 99°39'0″E), with the altitude of 2400-2800 m and annual temperature of 1.8 ℃. The annual precipitation of this region varies within a range of 270-350 mm, and the soil is mainly consisted of chestnut and chernozem soil. Xerophytic perennial grasses and mesohydriphytic bushes were the dominant native vegetation in this region. Variety of vegetation types lead to a rich biodiversity of grasshopper population by providing a heterogeneous habitat for the survival and reproduction of grasshopper. From June to August in 2009, 7 fixed observation plots (1hm²) along the elevation 50m were established. In each plot, 3-8 quadrates (20m×20m) were chosen according to typical location. In each quadrate, the pest-net (net diameter 30 cm) was used to seize grasshoppers by sweeping parallel 200 times, at the same time, vegetation community characteristics were investigated, including plant species richness, height, coverage and biomass. Thus the grasshopper species identification and quantitative statistics of grasshopper and plant variables were conducted in the laboratory. In this study, Detrended Correspondence Analysis (DCA) and Canonical Correspondence Analysis (CCA) were used to quantify the grasshopper population and analyze ecology relationship between grasshopper community spatial distribution pattern and vegetation community characteristic and domiant plant coverage. Besides, we analyzed the correlativity of grasshopper distibution number and dominant plant coverage. Results showed that a total of 1149 grasshoppers were collected, belonging to 3 families, 10 genera, 13 species. Oedaleus decoratus, Gomphocerus licenti and Chorthippus albonemus were the dominant species in this area. 13 species of grasshoppers were divided into six groups through DCA analysis. The number of grasshopper species, dominance, predatory behaviors and habitat choice of grasshopper are significantly different in those six groups. Meanwhile, no overlap was found within the same group in the hatch time, life form and tropic niche of grasshopper. The six grasshopper groups corresponded to different plant communities, and presented three kinds of correspondence relations, such as one-to-one correspondence relation between Chorthippus fallax and Stellera chamaejasme community; multiple-to-one correspondence of Oedaleus decorus, Myrmeleotettix palpalis and Filchnerella sunanensis with Artemisia dalailamae community; the rest of grasshopper community corresponded to multiple plant community types. There are three complex co-relationships between dominant plant coverage and grasshopper abundance, including extremely significant correlation, significant correlation and no correlation. Gramineae and Compositae plant have higher impact on the grasshopper distribution than other plants. Plant community affected grasshopper distribution pattern by providing food resource and proper habitat environment. The study of spatial distribution pattern and quantitative ecology relations between grasshoper and plant community and dominant plant, and the determination of the grasshoper species that most likely to have disaster in each plant community might be useful for predicting the grasshopper infestation and providing advices for grasshopper infestation control.