The invention discloses a compression method for test data of an irrational number storage test vector and relates to a fault coverage guided compression method for the test data of the irrational number storage test vector. The compression method comprises the following steps of firstly, generating a fault list according to a circuit structure of an integrated circuit to be tested; secondly, running an automatic test vector generation tool for faults to generate test vectors of corresponding faults; thirdly, counting the lengths of runs; fourthly, performing preliminary estimation on corresponding ranges of irrational numbers; fifthly, dichotomising the ranges of the irrational numbers, and successively approximating; sixthly, filling independent bits; seventhly, performing random test; eighthly, judging whether the fault list in the seventh step is empty or not, if the fault list is empty, turning to the ninth step, and otherwise, turning to the second step; ninthly, ending, and returning all records such as integers m and l corresponding to all the irrational numbers. According to the compression method disclosed by the invention, the coding of the irrational numbers and the generation of the automatic test vectors are combined, so that on one hand, coding numbers, corresponding to the test vectors, of easily-detected fault points are reduced, and on the other hand, the fault coverage is improved.