Managing water level from an ecological perspective has become an urgent issue in recent years in efforts to conserve and restore lake ecosystems. Here we provide a novel methodology to assess water level requirements (WLRs) of shallow lakes, combining geomorphological, hydrological and biological characters. The approach involves five calculation steps, and was applied in a shallow lake of the Yangtze River basin. Whole-lake aquatic vegetation coverage was regarded as a surrogate of ecosystem health. Water level and light availability, two major factors limiting the distribution of aquatic plants, were considered and quantitative relationships were established between water level during germination and vegetation coverage. The germination water level was then treated as a benchmark in determining WLRs of life history stages of aquatic plants. In the model, water levels during the early life history stages were held low and constant to enable germination, and the overall regime was matched with historical natural conditions. The case study showed that vegetation coverage decreased with increasing water level during germination, and a higher Secchi depth (greater water clarity) was associated with a larger coverage. The observed vs. estimated regression line was not significantly different from unity, indicating a high predictive power of the model. The methodology established a quantitative linkage between hydrological variables and ecosystem health. It could be widely used in WLRs assessments in Yangtze shallow lakes as well as other similar waterbodies, providing a useful tool to manage lake ecosystems for conservation and restoration. (C) 2017 Elsevier B.V. All rights reserved.