Available nitrogen is not only one of the most important limitation resources to manipulate primary productivity, but also a major factor to decide species composition in a grassland ecosystem. Chemolithotrophic nitrification is a two-step process, consisting of the conversion of ammonia to nitrite then further to nitrate. Those steps are involved two different groups of organisms, the ammonia-oxidizing bacteria (AOB) and the nitrite-oxidizing bacteria (NOB), respectively. In this study, the active AOB community was investigated by using approaches of polymerase chain reaction-denaturing gradient gel electrophoresis (PCR-DGGE) and sequences analysis on the Leymus chinensis steppes with three of mowing, grazing and fencing treatments respectively in Hulunbeier steppe of Inner Mongolia. Sampling in August 2008 when vegetation was at its best condition. Samples were taken at depths of 0-0.20m. The community structure and diversity of the bacterial groups from the different samples was further analyzed by using different techniques, such as statistical analysis and diversity index evaluation of the band patterns etc. The results showed that grazing activity significantly increased the number of species and quantities of AOB, as well as the diversity. A total of 21 sequences representing 21 different sequence types were recovered from the DGGE gels after phylogenetic reconstructions. The results further revealed that most sequences were coming from Nitrosospira and Nitrosomonas of β-Proteobacteria, with accounting for 72.4% and 28.6% respectively. Cluster 4 of Nitrosospira existed in grazing, mowing and fencing plots with accounting for 25.0%, 27.2% and 26.3% respectively in their dominant AOB. And Cluster 6 of Nitrosomonas existed in grazing, mowing and fencing plots, those accounted for 33.1%, 18.1% and 21.1% respectively in their dominant AOB. Analyze of stepwise regression showed that the key soil physicochemical factor which influence the diversity index most is the content of nitrate nitrogen. Urease activity and nitrate rate in grazing plot were both higher than the other land use patterns. The content of nitrate nitrogen and total phosphorus were both extremely significant (P<0.01) correlated with urease activity, and there is a significant (P<0.05) relationship between urease activity and the content of available phosphorus; at the same time, there is also a significant relationship between nitrate rate and nitrate nitrogen contents. AOB communities in Leymus chinensis steppe were significantly (P<0.05) influenced by the levels of ammonium nitrogen when canonical correspondence analysis was employed to identify relationship between AOB and soil physicochemical factors under different land use patterns. Based on the above findings, it seems that nitrogen cycles represent the key links between above and below ground ecosystems. Land use patterns in Leymus chinensis steppe have significant and long-term impacts on the size and the structure of the soil AOB community at both phylogenetic and functional levels, thus would further affect the transformation and utilization of nitrogen in the steppe ecosystem.