CONTEXT: Crop-livestock systems in China have become the primary agricultural nitrogen (N) pollution source owing to fertilizer overuse and poor manure management. Yet city-scale analyses of nitrogen use efficiency (NUE) trends and loss hotspots remain insufficient. Current evaluations fail to fully account for spatial variations in farming practices, natural conditions, and sector-specific contributions to N losses. OBJECTIVE: This study aims to quantify N flows, reveal NUE evolution, identify hotspots of N losses, and evaluate mitigation potentials and optimization paths for regionally differentiated N management. METHODS: We improved the Nutrient Flows in Food chains, Environment, and Resources use (NUFER) model by innovatively introducing human excreta indicators and livestock localization parameters, establishing a city-scale comprehensive framework. RESULTS AND CONCLUSIONS: The system efficiencies showed marked improvement. From 2000 to 2022, NUE showed a relative increase of 39.91% (crop systems), 46.30% (livestock systems), and 43.47% (crop-livestock systems) compared to the 2000 levels, with pronounced spatial variation. High-performing crop systems (NUE > 50%) clustered in the Northeast China Plain and Huang-Huai-Hai Plain (HHHP), while inefficient livestock operations (NUE < 15%) predominated in western China and the Yunnan-Guizhou Plateau (YGP). Notably, fruits and vegetables (cash crops) showed significantly lower NUE than grain crops. Maize, vegetables, rice, fruits, and wheat contributed 82.29% of crop N losses, while beef cattle, pigs, and poultry accounted for 83.52% of livestock losses. Notably, fruits and dairy cattle showed the highest N loss intensities at 110.07 kgha(-1) and 31.72 kglu(-1), respectively. While NH3 dominated most regions, runoff and erosion prevailed in the Sichuan Basin and Surrounding Regions (SBSR) and YGP. We identified concentrated hotspot regions - representing just 6.17% of Chinese land area but contributing 30.42% of national N losses - primarily located in: (i) the central and southern HHHP, (ii) the northern Middle-Lower Yangtze River Regions, and (iii) Central SBSR, Loess Plateau, and YGP. Scenario analysis demonstrated substantial mitigation potential. Integrated measures (precision fertilization, resource recycling, livestock structure optimization, and advanced manure management) could reduce synthetic N inputs by 22.64% while decreasing N losses by 5.03 Tg (31.38%). SIGNIFICANCE: This study provides a scientific basis for spatially differentiated N management strategies in both Chinese and other countries' worldwide agricultural systems.