Abstract: Abstract：In order to evaluate the health conditions of rivers and solve the problem of the highly subjective weights of evaluation indicators in river health assessment, BP neural network is used to simulate respectively the ecological status and social service function of the Yuan River and to judge by coordination degree whether they are developing in a coordinative way. The results show that: (1) The comprehensive evaluation index of the river health within Yichun section and Xinyu section in 2016 is greater than 0.6, which indicates that the river is healthy. The coordination index of Yichun section is 0.8, which shows it is well coordinated. In comparison, the coordination index of Xinyu section is 0.79, which means it is moderately coordinated. (2) Compared with the previous year, the natural ecological status index  fx of Yichun section and Xinyu section in 2006 and 2010 decreased, comparative indicator data ,which was mainly caused by the degraded water quality of the Yuan River that went down from the previous Grade III to the present Grade IV;(3) The natural ecological status index fx of Xinyu section in 2016 stands at 0.59, which is much smaller than the social service function index of 0.68. Therefore, it is suggested that the natural ecological restoration be privileged in the future river policy for Xinyu section of the Yuan River.