The invention relates to a high-speed rail power quality abnormal condition prewarning method based on a dynamic threshold. The method includes the following steps of (1) building a high-speed rail power quality prewarning index system, (2) collecting historical power quality data and traction load amount corresponding to the historical power quality data, (3) building a high-speed traction station power quality abnormal state prewarning reference database, (4) detecting a power quality prewarning index abnormal state during online monitoring and prewarning, and (5) prewarning the high-speed rail power quality abnormal state. By means of the historical power quality monitoring data, the high-speed rail traction station power quality abnormal state which is monitored in real time is prewarned, power supply and consumption parties can learn about the high-speed rail traction station power quality abnormal state, power quality potential risks can be found in advance, operation reliability and economical performance of a power grid are improved, and safe operation of the power grid is guaranteed.