Nitrogen pollution is a serious environmental issue in the Danjiangkou Reservoir region (DRR), the water source of the South-to-North Water Diversion Project of China. In this research, seasonal surveys and a bi-weekly time series survey were conducted in the Qihe River Basin, one of the most densely populated agricultural basins in the DRR. Hydrochemical compositions (NO3− and Cl−), dual isotopes (δD-H2O, δ18O-H2O, δ15N-NO3−, and δ18O-NO3−), and a Markov Chain Monte Carlo isotope mixing model were jointly applied to unravel the sources, migrations, and transformations of the nitrate (NO3−) in the basin. It was revealed that the mixing between different sources was the main process controlling the isotopic compositions of the riverine NO3− in the upper-middle reaches. In contrast, denitrification occurred in the lower reaches. For the first time, the sources of NO3− were quantified at a basin scale in the DRR. Overall, the river transported 484.2 tons/year of NO3-N to the reservoir, of which 32.6%, 36.4%, 28.0%, and 3.0% was from soil organic nitrogen, chemical fertilizer, residential sewage and atmospheric precipitation, respectively. The NO3-N fluxes of the different sources were regulated by the monsoon climate and anthropogenic activities. For example, high precipitation and intense fertilization resulted in severe nonpoint source pollution. Denitrification thrived in soils and reservoirs in wet seasons. Temperature could regulate the migration, nitrification and denitrification processes. Based on the results, we suggest that the management strategies dealing with nitrogen pollution issue in the DRR should follow the specific spatiotemporal characteristics of NO3− sources, migration and transformation mechanisms.