Debris flow is a kind of serious geologic hazard and Real-time prediction of debris flow is essential for the protection of human life and property. The objective of this paper is to propose a method for real-time predicting debris flow using fuzzy event tree (FET), which is constructed by integrating event tree analysis, fuzzy mathematics and expert knowledge. Xiu Yan County, An Shan City, Liaoning province in China was used as a case. The prediction assessment was based on four dynamic process factors, namely antecedent rainfall (AR), the maximum rainfall intensity (MRI) each hour, soil saturated liquefaction degree (SSLD) and soil instability (SI). The four dynamic process factors were viewed as initiating events in FET model for forecasting debris flow. The result indicated that the majority of study area is characterized by medium and low dangerous level. This study suggests that weather forecast, timely detection of SSLD and SI are important for preventing and mitigating disaster and FET methodology is promising for real-time foresting geologic hazard.