Identifying floodplains with high rates of denitrification will help prioritize restoration projects for the removal of nitrogen. Currently, relationships of denitrification with hydrogeomorphic, physiographic, and climate (i.e., largescale) characteristics of floodplains are relatively unknown, even though these characteristics have datasets (e.g., geographic mapping tools) that are publicly available (or soon-to-become) that could be used to understand denitrification variability. Thus, we investigated control of denitrification by these largescale characteristics in eighteen nontidal floodplains of the Chesapeake Bay watershed (i.e., at regional scale, >100 km, scale), using measurements or compiled data at the scales of the stream reach and respective catchment; floodplain soil and herbaceous vegetation (i.e., local) characteristics were additionally investigated. Soil denitrification potentials were measured in May, July, and August using complementary acetylene-based techniques under an anoxic environment. Linear largescale predictors of denitrification potential measurements included stream nitrogen and phosphorus concentrations (+), channel width-to-depth ratio (+), floodplain sedimentation (+), forested (−) and urban (+) catchment land cover, and seasonal air temperature (−). Three predictors, catchment forested land cover (strongly related to agricultural land cover), catchment urban land cover, and floodplain sedimentation were related to the most number of denitrification potential measurements. Soil structure, soil nutrient concentrations, and herbaceous vegetation characteristics that were seasonally measured (with a few exceptions) were linear predictors of denitrification potentials in May and August, with nitrogen and carbon characteristics the most consistent (positive) predictors across measurements. Nutrient amendment assays further supported the importance of nitrogen and carbon controls. Using the local characteristics as statistical mediators in path analysis, greater non-forested catchment land cover indirectly increased denitrification through greater floodplain soil nitrate, total phosphorus, and herbaceous aboveground biomass. Additionally, greater floodplain sedimentation indirectly increased denitrification through greater soil pH, total phosphorus, and potential carbon mineralization. Due to the consistency of relationships across denitrification potential measurements along with path modeling results, hotspots of floodplain denitrification should be found in urban and agricultural catchments where river-floodplain hydrologic connectivity promotes sedimentation. Largescale predictors explained 43–57% of the variation in denitrification potentials and should be useful for prediction in floodplains. Siting restoration projects in watersheds for maximum nitrate removal using publicly available largescale datasets is both feasible and effective.