The emission factor (EF₅) approach is widely used to estimate N₂O emissions from rivers and estuaries. However, current IPCC default EF₅ values use fixed and identical factors for both systems (0.26%), without accounting for environmental variability or the substantial differences between these two systems. Here, we investigated EF₅ variability and its controls along the urbanized Pearl River-estuary continuum using isotopocule analysis, metagenomics, and global data synthesis. By integrating our field measurements with compiled global datasets, we found that EF₅ ranged from 0.0014% to 18.6% in rivers (n = 3,560) and from 0.01% to 12.81% in estuaries (n = 254). Across the continuum, isotopocule analysis suggested that denitrification dominated N₂O production (>89%), while SHAP analysis identified nitrate as the strongest EF₅ predictor. Lower EF₅ under nitrate-rich conditions was associated with uneven denitrifier responses: higher nitrate increased nirS but not nirK abundance and was more strongly associated with upstream nitrate-reduction genes than with downstream N₂O-producing genes. Phosphorus-related variables also explained substantial EF₅ variation. Higher total phosphorus concentrations were associated with lower EF₅ and greater potentials for assimilatory nitrate uptake and DNRA, indicating that more nitrate may have been consumed by non-N₂O-producing pathways. Predictor importance was zone dependent: nitrate-related factors were more important in low-nitrate estuaries, whereas phosphorus-related factors became primary factors in rivers. In urbanized riverine reaches, TN:TP was positively associated with EF₅, partly reflecting wastewater influence, where elevated TN:TP coincided with disproportionately greater dissolved N₂O inputs relative to nitrate. Global synthesis further showed that estuarine EF₅ was generally higher than riverine EF₅, with high salinity inhibiting N₂O reduction. These findings demonstrate that riverine and estuarine EF₅ differ in both magnitude and dominant controls, providing a basis for system-specific prediction and refinement of aquatic N₂O emission factors.