Transient safety assessment of hydroelectric generation systems is a major challenge for engineers specialized in hydropower stations around the world. This includes two key scientific issues: the dynamic risk quantification in a multi-factors coupling process, and the identification of elements with highest contribution to system stability. This paper presents a novel and efficient dynamic safety assessment methodology for hydroelectric generation systems (HGSs). Based on a comprehensive fuzzy-entropy evaluation method, the dynamic safety level of the system is estimated by means of probability values, and the influence rate of assessment indices on the HGS risk profile is also obtained. Moreover, a number of risk mitigation and maintenance amendment strategies are discussed to reduce the losses in operation and maintenance (O&M) costs at hydropower stations. The methodology is implemented and validated using an existing hydropower station experiencing a start-up transient process, results of which are shown to be beneficial to operators and risk managers. It is recommended that the presented methodology is applicable not only to the HGS's start-up process but is also promisingly useful for largely fluctuating transient processes of other engineering facilities.