Riparian lowlands are located at the interface between streams and upland areas. Water fluxes from surrounding areas must pass through these areas to reach the stream, and riparian lowlands may thus have a great influence on stream water quality. For this reason, riparian lowlands have been intensely studied during the past 30 years. However, the vast majority of these studies have been conducted in sandy settings, while studies in glacial till landscapes are scarce. This study identified flow paths and quantified water and nitrogen (N) fluxes through four subareas of a 26 ha Danish riparian lowland in a Weichselian glacial till landscape. These subareas were all located at major drain outlets where drain water from upland areas discharged onto the surface of the riparian lowland in the hillslope at the upland-riparian lowland border. A water balance was set up for each of the subareas by continuous monitoring of incoming drain water, precipitation, and evapotranspiration, and discrete measurements of groundwater flow. Groundwater flow entering and leaving the riparian lowland was calculated for hydrostratigraphic layers delineated from an extensive number of boreholes and geophysical data. These groundwater fluxes were calculated as Darcy fluxes from hydraulic heads and hydraulic conductivities measured in piezometers installed at several depths along profiles parallel and orthogonal to topographical gradients. Unmonitored fluxes included overland flow to the stream and subsurface drain flow to the stream from drain pipes within the riparian lowland. These fluxes were modelled as residual fluxes after taking storage within the riparian lowland into account. Concentrations of N species were measured in water samples from these flow paths, which enabled the calculation of N fluxes. Overland flow was the major flow path along which water entered the stream from the riparian lowland in all the investigated areas. However, groundwater and subsurface drain flow also contributed substantially in some of these areas, demonstrating the heterogeneity in the flow path distribution even within a small riparian lowland. The flow path distribution was controlled by the hydraulic loading rate (HLR) to the riparian lowland and the hydraulic properties of the riparian lowland soils. Incoming water in excess of the infiltration and/or drainage capacity of the riparian lowland soils would travel to the stream as direct overland flow. Thus, the relative contribution of overland flow to the stream was determined by both HLR and the hydraulic properties (presence of tile drains, hydraulic conductivities, hydraulic gradients, and thickness of permeable sediments) of the riparian lowland soils. The major N inputs to the riparian lowland was in the form of nitrate (NO3-) from the upland drains. In one of the investigated subareas, the riparian lowland was subject to a high HLR, the distance from the hillslope drain was short, and the terrain slope in the riparian lowland was steep. In this area incoming drain water was allowed to travel unaltered as direct overland flow transporting NO3- to the stream. In this area the removal of NO3- was 25%, while the overall removal efficiency of total nitrogen (TN) was 1%. In the remaining investigated areas, overland flow to the stream was also the dominant flow path. However, the majority of this overland flow was either i) allowed to infiltrate in the upslope parts of the riparian lowland before exfiltrating in the downslope parts (short return flow) ensuring interaction with the riparian lowland soils, and/or ii) sufficiently dispersed to allow exchange of NO3- from surface water to the underlying sediment by diffusion. Removal of NO3- in these areas was 71-94%, while the overall efficiency for TN was 39-56%. The discrepancy between removal efficiencies of NO3- and TN was due to a substantial export of organic N (Norg), mainly via overland flow, and to a smaller degree export of ammonium (NH4+) via groundwater, and both Norg and NH4+ from subsurface tile drains. Thus, while the presence of subsurface tile drains within the riparian lowland reduced overland flow and thereby export of NO3- and Norg, this was counteracted by transport of Norg and NH4+ via subsurface tile drain discharge. Concurrent monitoring of discharge and concentrations of N-species in the stream at the catchment outlet confirmed that Norg was the main export from the entire catchment, and both the dynamics of N fluxes and the total annual N fluxes could be estimated from a simple area- or flow-upscaling of the investigated subareas. Export of Norg from riparian lowland soils was estimated to contribute ~40% (21-49%) of total catchment N-transport. Removal of NO3- in wetlands has mainly been attributed to heterotrophic denitrification and this is often the only N-removal pathway included in operational N-balance models. An increasing number of studies have reported a significant importance of other N-transport pathways, such as dissimilatory nitrate reduction to ammonium (DNRA) or anaerobic ammonium oxidation (anammox). At a specific site, where NO3- was seen to infiltrate the riparian lowland soils whereafter it disappeared, the relative importance of different N-transformation processes was investigated using an isotope pairing technique by amending soil slurries with combinations of 15N-labelled NO3- and NH4+. Results showed that heterotrophic denitrification to N2 was responsible for >78% of NO3- removal in slurry incubations at 20°C, while anammox and DNRA contributed <8% and <5%, respectively. In slurries amended with NO3- with no addition of Fe2+, 2-11% of the transformed NO3- was converted to N2O. In slurries with both added NO3- and Fe2+, N2O production increased by a factor 10-70, corresponding to a conversion of 43-75% of the transformed NO3- to N2O.