A framework called Object-Oriented Precise Decision-making (OOPD) was proposed which oriented to the lake itself. The framework was based on Numerical Source Apportionment powered by 3-dimensional water quality model, which then quantified the causality of water quality improvement and load reduction. The proposed framework was applied to support short-term decision making of Lake Yilong, a eutrophic lake. Results showed that Chenghe sub-watershed and Chengbeihe sub-watershed were important pollution source no matter considering which monitoring station and under which water diversion scenario. In addition, comprehensive pollution control should be considered to ensure that Huzhong monitoring station or Hudong monitoring station meet water quality standards. However, considerable load reduction cannot guarantee water quality if there was no water diversion. Water quality of three monitoring stations would be improved a lot under 20 and 30 million m³ annual water diversion scenario. Finally, based on the analyses above, suggestion of focused pollution control project was given for each sub-watershed and an evaluation of one bean product wastewater treatment extension project was given to illustrate how to combine micro and macro aspects in OOPD.