Abstract: With the initial achievements in basin ecological protection, the advancement of green basins has gained increasing attention. Establishing a rational evaluation indicator system is essential for supporting the planning and implementation of green basin initiatives. This study proposes a method for selecting green basin indicators based on a Binary Integer Programming (BIP) model, aiming to optimize the construction of a green basin indicator system. The methodology includes: ① constructing a preliminary evaluation indicator system and eliminating indicators with significant data gaps through data analysis to improve data quality; ② determining the weight of each indicator using a combined entropy weight and CRITIC method and excluding indicators with extremely low weights to enhance the representativeness of the indicator system; ③ formulating a BIP model and applying a genetic algorithm to solve for the optimal indicator set. A evaluation approach of "single indicator quantification - multi-indicator synthesis - multi-criteria integration" is adopted to validate the optimized indicator system. An empirical study is conducted in the Changjiang River Basin. The results show that the proposed method can effectively identify 25 key indicators closely associated with green basin performance. The research findings provide an effective indicator selection and optimization approach for the scientific management of green basins.