Transient safety assessment of hydroelectric generation systems is a major challenge for engineers specializing in hydropower stations worldwide. This includes two key scientific issues: the dynamic risk quantification in multi-factors coupling process, and the indices identification with high contributions on system stability. This paper presents a new flexible, rapid and affordable dynamic safety assessment methodology for a hydroelectric generation system. Based on the fuzzy-entropy comprehensive evaluation method, the dynamic safety level of the system is estimated by means of probability and the influence contributions of assessment indices on risk operations of the hydroelectric generation system are also obtained. Moreover, some risk mitigation and maintenance strategies are finally discussed to reduce dual losses of operation and maintenance in hydropower stations. The methodology is implemented and validated in an existing hydropower station aiming at a start-up transient process, which is beneficial to risk warning and maintenance strategy enhancement. In addition the presented methodology in this paper is not only applied in the start-up transient process 2 but is also promisingly and appropriate for other large fluctuation transient processes.