The invention discloses a mass data-based power system safety and stability judgment method taking topological factors into consideration. The method comprises the following steps: (S1) determining active power Pi and reactive power Qi, which flow through various lines in a power system; (S2) building power grid operation mode description taking topological factors into consideration; (S3) judging whether the power system runs safely and stably; (4) repeating the steps (S1) to (S3), accumulating for a period of time to obtain m groups of samples, which include operation modes x (j=1, 2, ..., m) of the power system and a tag y (j=1, 2, ..., m) that the power system can run safely and stably in the operation mode; (S5) building a nonlinear support vector machine model (SVM) by m groups of samples obtained in the step (S4) and finding out an interface f(x) that the power system can run safely and stably and cannot run safely and stably; and (S6) judging whether the power system can run safely and stably or not according to the online operation state of the power system by the nonlinear support vector machine model (SVM). The mass data-based power system safety and stability judgment method taking topological factors into consideration is high in response speed in judgment and applicable to power system safety and stability judgment in a complicated variable online power grid operation mode.