The southwestern Adirondack region of New York receives among the highest rates of atmospheric nitrogen (N) deposition in the USA. Atmospheric N deposition to sensitive ecosystems, like the Adirondacks, may increase the acidification of soils through losses of exchangeable nutrient cations, and the acidification of surface waters associated with enhanced mobility of nitrate (NO3−). However, watershed attributes, including surficial terrestrial characteristics, in-lake processing, and geological settings, have been found to complicate the relationships between atmospheric N deposition and N drainage losses. We studied two lake-watersheds in the southwestern Adirondacks, Grass Pond and Constable Pond, which are located in close proximity (∼26 km) and receive similarly high N deposition, but have contrasting watershed attributes (e.g. wetland area, geological settings). Since the difference in the influence of N deposition was minimal, we were able to examine both within- and between-watershed influences of land cover, the contribution of glacial till groundwater inputs, and in-lake processes on surface water chemistry with particular emphasis on N solutes and dissolved organic carbon (DOC). Monthly samples at seven inlets and one outlet of each lake were collected from May to October in 1999 and 2000. The concentrations of NO3− were high at the Grass Pond inlets, especially at two inlets, and NO3− was the major N solute at the Grass Pond inlets. The concentrations of likely weathering products (i.e. dissolved Si, Ca2+, Mg2+, Na+) as well as acid neutralizing capacity and pH values, were also particularly high at those two Grass Pond inlets, suggesting a large contribution of groundwater inputs. Dissolved organic N (DON) was the major N solute at the Constable Pond inlets. The higher concentrations of DON and DOC at the Constable Pond inlets were attributed to a large wetland area in the watershed. The DOC/DON ratios were also higher at the Constable Pond inlets, possibly due to a larger proportion of coniferous forest area. Although DON and DOC were strongly related, the stronger relationship of the proportion of wetland area with DOC suggests that additional factors regulate DON. The aggregated representation of watershed physical features (i.e. elevation, watershed area, mean topographic index, hypsometric-analysis index) was not clearly related to the lake N and DOC chemistry. Despite distinctive differences in inlet N chemistry, NO3− and DON concentrations at the outlets of the two lakes were similar. The lower DOC/DON ratios at the lake outlets and at the inlets having upstream ponds suggest the importance of N processing and organic N sources within the lakes. Although an inverse relationship between NO3− and DOC/DON has been suggested to be indicative of a N deposition gradient, the existence of this relationship for sites that receive similar atmospheric N deposition suggest that the relationship between NO3− and the DOC/DON ratio is derived from environmental and physical factors. Our results suggest that, despite similar wet N deposition at the two watershed sites, N solutes entering lakes were strongly affected by hydrology associated with groundwater contribution and the presence of wetlands, whereas N solutes leaving lakes were strongly influenced by in-lake processing. Copyright © 2006 John Wiley & Sons, Ltd.