The invention relates to a rail transport passenger flow prediction method based on a rail IC card and mobile phone signaling data. The method comprises the following steps of S1, collecting rail transport passenger IC card data information and carrying out OD matching on pretreatment data; S2, carrying out commuting identification, inputting data of working days, grouping according to users, and determining a commuting rule of each user in the working days: when same inbound site and outbound site days of the user reach 50% of total days, determining that a trip is a commuting trip of the user; S3, calculating a rail transport passenger flow, wherein statistics indexes include a site passenger flow, a line passenger flow, a section passenger flow, a transfer passenger flow, a passenger flow OD, a commuting site passenger flow, a commuting line passenger flow, a commuting site OD passenger flow and a commuting line OD passenger flow; S4, carrying out normal state passenger flow volume analysis and holiday passenger flow volume analysis; and S5, carrying out normal state passenger flow volume prediction and holiday passenger flow volume prediction. By using the method, a rail transport passenger flow volume can be accurately predicted, and the method has an important effect on a reasonable compiling transport power configuration scheme, a passenger flow organization scheme and the like.