© 2019 Elsevier B.V. Globally, small agricultural waterways fed by springs, tile drains, and seeps can disproportionately contribute to downstream nutrient loading, which is associated with declines in water quality and ecosystem functions. Treating nitrate using a multiple tool, multiple-scale approach in small waterways could offer improved management of these sources. We used a before-after-control-impact design to test the suitability of three small (<30 m 3 ) edge-of-field denitrifying woodchip bioreactors and stream bank re-shaping and riparian planting. Over three-and-a-half-years, riparian rehabilitation enhanced nitrate flux attenuation compared to pre-rehabilitation, but only under relatively low flow conditions. In comparison, there were no significant changes in nitrate flux in a control waterway under any flow condition. N fluxes always increased in both the control and treatment waterways when reaches gained water downstream. Nitrate removal efficiencies for all three bioreactors ranged from <10 to >99%, with performance variations due to short residence times and fluctuations in source water chemistry. A single tile drain bioreactor removed 0.41 kg NO 3 -N d −1 , equivalent to ∼10% of the mean daily tile drain nitrate load. Greenhouse gas fluxes from the tile drain bioreactor were similar to the surrounding pasture (CO 2 -C mean: 185–286 mg C m 2 h −1 ; N 2 O-N mean: 49–90 μg N m 2 h −1 ), suggesting no negative impacts from the bioreactor. Overall, our results suggest a multiple-tool, multiple-scale application of rehabilitation tools can reduce downstream N fluxes, but only under certain flow conditions. Thus, local rehabilitation tools, like those trialed here, will need to be scaled appropriately if they are to significantly attenuate nutrient losses from small agricultural waterways. Moreover, these will not replace catchment-scale nutrient plans to address losses from land and legacy groundwater N pollution.