Abstract: Given the significant water level fluctuations in river-connected lakes and the distinct habitat functions across different seasons, the existing methods predominantly focus on static water levels or individual species, making it challenging to simultaneously address dynamic hydrological changes and multi-habitat requirements. This study proposes a novel method for determining the ecological water level of river-connected lakes by integrating the habitat suitability of multiple species. Membership functions are employed to quantitatively characterize the water level requirements of various key species during their critical growth stages within the lake area. A game theory-based weighting approach is then introduced to reconcile potential conflicts among the water level demands of different key species. Combined membership functions for each time period are calculated using the derived weights, and the water level range corresponding to the maximum combined membership degree is identified as the ecological water level range for that period. Taking Shijiu Lake, a river-connected lake in the lower reaches of the Yangtze River, as a case study, the proposed method was applied to determine the suitable ecological water level ranges from January to December. The results indicate that the daily water level rise during May—June should be approximately 0.02 m/d, while during the transition period (3—8 days) from June to July, the daily rise should be between 0.3—0.5 m. The findings provide a quantitative basis for the dynamic water level management of Shijiu Lake, and also offer a novel approach for determining ecological water levels in other river-connected lakes.