Nitrogen deposition is changing biodiversity on Earth. We need to understand the underlying mechanisms to conserve biodiversity better. Both selection and chance are potential mechanisms, and they may operate concurrently. Then, what are the respective effects of selection and chance, what is their relative importance and how do they change with increasing nitrogen deposition rate? Here, we performed a 6-year nitrogen addition experiment (0-28 g N/m2/year) in a typical steppe ecosystem of Inner Mongolia to investigate the community structure of plants, bacteria and ammonia-oxidizing Archaea (AOA). We developed an experimentally based calculation method to first separate the structural variations between plots into the effects of selection (S) and chance (C), and then calculate their relative importance. Our results showed that as nitrogen addition rate increased, S for both plants and bacteria increased, but C for plants first increased and then decreased, and C for bacteria also increased; meanwhile, both S and C for AOA changed nonlinearly. As nitrogen addition rate increased, the importance of chance decreased on the whole for all these communities, but it decreased nonlinearly for plants and bacteria, with a local increase at certain intermediate rates. At all treatments, the importance of chance was < 0.5 for plants, but > 0.5 for AOA. These results demonstrated that nitrogen deposition changed biodiversity by mediating the effects and importance of chance, implicating different strategies should be adopted in conserving biodiversity according to nitrogen deposition rate and community properties.