Nitrogen (N) deposition in many areas of the world is over an order of magnitude greater than it would be in absence of human activity. We ask how abiotic (N)and biotic (plant host and neighborhood) effects interact to influence root-associated bacterial (RAB)community assembly. Using 454 pyrosequencing, we examined RAB communities from two dominantal pine tundra plants, Geum rossii and Deschampsia cespitosa, under control, N addition and D. cespitosa removal treatments, implemented in a factorial design. We hypothesized that host would have the strongest effect on RAB assembly, followed by N,then neighbor effects.The most dominant phyla were Proteobacteria (mostly Gammaproteobacteria), Actinobacteria,Bacteroidetes and Acidobacteria. We found RAB communities were host specific, with only 17% overlap in operational taxonomic units. Host effects on composition were over twice as strong as Neffects. D. cespitosa RAB diversity declined with N, while G. rossii RAB did not. D. cespitosa removal did not influence G. rossii RAB community composition, but G. rossii RAB diversity declined with N only when D. cespitosa was absent. We conclude that RAB of both hosts are sensitive to N enrichment, and RAB response to N is influenced by host identity and plant neighborhood.