The invention a method for comprehensively predicting power transmission line fault rate based on multidimensional cloud model. According to calendar year statistical data for power transmission line fault rate under the extreme external environments such as lightning, icing, and pollution flashover, correlation of the external environments such as lightning, icing and pollution flashover is mined by using big data mining technology, and statistical data for power transmission line fault rate under the various external environments not correlated is acquired by using Cholesky decomposition process; a multidimensional cloud estimation model for calendar year statistical data under various extreme external environment factors and a one-dimensional cloud model for power transmission line fault rate are acquired by using one-dimensional reverse cloud generator generating algorithm; the fault rate model for power transmission lines under various external environments is considered herein, the fault rate of a power transmission line can be predicted quantitatively, prediction is highly accurate, calculation is convenient, and the method is well applicable to the prediction of fault rate of large grid transmission lines.