Water quality in shallow Lake Taihu, China, has rapidly deteriorated in the past decades due to rapid urbanization and changed land use, resulting in severe blooms of toxic cyonabacteria (figure 1) and reduction of submerged macrophytes. Climate change has intensified these problems. This study on Lake Taihu has six key objectives: 1) assess current ecosystem health; 2) identify the desired ecological quality; 3) identify critical levels of nutrient loading necessary to reach this quality; 4) identify the uncertainty in these critical levels; and 5) evaluate scenarios that mitigate eutrophication and other major stressors on ecosystem health. We want to identify the critical nutrient loadings for Lake Taihu with a model for integrated water system research. These critical nutrient loadings define the points at which the lake suddenly switches from a good ecological quality into a bad ecological quality and vice versa. In Europe, knowledge of critical loadings has become very important due to the EU Water Framework Directive that imposes high quality standards for all water bodies in 2015. The focus on critical nutrient loading is original and innovative. While the concept of nutrient loading as such is key to any limnological study on lake eutrophication, the concept of critical nutrient loading has only emerged when it became apparent that ecosystems often respond in a non-linear way to external stressors. A key innovative element in the topic of this project is therefore that we study nutrient loadings in the light the positive feedbacks that maintain alternative stable ecosystem states. We will use the existing ecosystem model PCLake as a template for our ecosystem model of Lake Taihu. PCLake has been developed, tested and implemented as an important tool for managing water quality in shallow lakes in the Netherlands and a formal protocol for analyzing uncertainty in model output is available.