The Biliuhe reservoir, which is one of the most important water sources in Dalian, is the "source of life" of Dalian. To explore the characteristics of heavy metal pollution in Biliuhe reservoir, the potential ecological risk of these heavy metal elements were evaluated by two methods, i.e. geo-accumulation index and potential ecological risk index, and the risk values of heavy metals in sediment were predicted one year later in the future by grey forecasting model GM (1, 1) and RBF neural network. The comprehensive assessment results showed that the pollution level in the reservoir was slight, Hg was the primary element with higher ecological risk while Cu was secondary. The forecasting results indicated that the predicted risk value of Hg was higher than others. The results were discussed in order to provide the theoretical basis for the management of the reservoir in the future. The models will be used widely in the field of ecological risk assessment of heavy metals in sediment, due to the simplification and maneuverability of them.