Abstract: Collaborative environmental protection and food security are of great significance for high-quality agricultural development in lake basins. Rice, the primary grain crop in China, is widely planted in lake basins. However, previous studies on nitrogen (N) in rice across basins often ignored the spatial heterogeneity of emission factors and impact of field-integrated technological optimization on rice yield increases, while emission reductions remain unclear. This study aimed to clarify the current status and reduction potential of N fertilizer of rice planting system in the Erhai Basin using survey data from 194 households and 322 literature sources on reactive nitrogen loss. Life cycle assessment and random forest models were employed for comprehensive evaluation. The emission reduction and increased production potential of this basin were predicted based on 836 site-year field experiments conducted in the southwest region. Results displayed that average rice yield was 8598.5 kg·hm⁻², with an average N fertilizer input of 222.0 kg(N)∙hm⁻². The types of N fertilizer used included chemical and organic sources, with an average application of 95.6 kg(N)∙hm⁻² for chemical N and 126.4 kg(N)∙hm⁻² for organic N, in which the organic N accounted for 56.9% of total N fertilizer application. The regions with high N input and surplus areas were mainly located in the north, including Zibihu, Dengchuan, Fengyu, and Yousuo towns, as well as in the western part, including Shangguan, Xizhou, Wanqiao, Yinqiao towns, and Haidong Town in the eastern part. The average N losses was 55.1 kg(N)∙hm⁻², with nitrous oxide emissions, ammonia volatilization, surface N runoff, and N leaching accounted for 0.8%, 61.3%, 15.1% and 22.9%, respectively. The spatial distribution characteristics indicated a higher environmental risk in the northern and western part. Combined with the rice-planting region area, the total N loss was 440.0 t in the Erhai Lake Basin. The random forest model predicted that the rice production system could reduce reactive N losses by 22.9% while increasing rice yield by 21.1%. Compared to traditional farming, optimized technology can reduce N input by 428.5 t and reduce N losses by 105.6 t. By implementing optimized technology, the rice planting system in the Erhai Lake Basin could achieve increased yield, and reduced emissions. The study results provide a good reference for the green and high-quality development of plateau lake basins agriculture.