Abstract: LID (low impact development) facilities are important in regulating extreme rainfall, resisting flood disasters, improving construction of “sponge cities” . Design of LID facilities, however, often ignores their ecosystem environmental benefits, there is also some lack of evaluation methods for comprehensive benefits of these facilities． Realizing the layout and optimization of LID facilities at small and medium scales and incorporating various added values of LID facilities into comprehensive evaluation system has become a research hotspot in this field． Taking a built-up community in Beijing as an example, this paper constructs a stormwater management model using data from terrain and rainwater pipe network． The benefits of LID facilities on stormwater reduction are evaluated, and the ecosystem service value provided by these facilities is quantified using a monetization method, including initial investment cost, maintenance cost and ecosystem service value provided by these facilities during their life cycle． Stormwater management model is combined with benefit evaluation to form a coupled optimization model． Considering growth potential of maintenance cost and ecosystem service value of LID facilities in their life cycle, the spatial layout of LID facilities was optimized by non-dominated sorting genetic algorithm (NSGA-Ⅱ)． Facility area was used as decision variable, runoff reduction rate and net cost were used as objective function． Entropy weight-TOPSIS method was used to comprehensively evaluate and optimize the scheme． Under the premise of meeting relevant constraints, a runoff reduction rate of 76.3% was achieved, comprehensive net cost was 1.4×10⁶ yuan． This study aims to provide more comprehensive decision-making for sponge city construction, to promote reuse of water resources and improvement of urban environment, to provide important urban ecological support and emission reduction channels for achieving carbon peak and carbon neutrality goals by evaluating comprehensive benefits of LID facilities． The algorithm and monetization results of added value of facilities in this study have a certain degree of universality． After appropriate adjustment of parameters according to actual engineering, they can be applied to other study areas, providing reference for optimized layout of LID．