A method and apparatus for ranking combinations of mutants, test cases and random seeds in mutation testing, comprising obtaining, based on a signal of a test case target, logic gates related to the signal of the test case target and mutants on the related logic gates, for a compiled integrated circuit under test; calculating distances between the mutants and the signal of the test case target; performing a circuit simulation on the compiled integrated circuit under test to obtain activation cycle numbers corresponding to combinations of the mutants, test cases and random seeds; obtaining activation cycle number variances corresponding to combinations of the mutants and the test cases; and ranking the combinations of the mutants, the test cases and the random seeds based on the distances, the activation cycle numbers and the activation cycle number variances. The invention can reduce the probability that the mutation simulation selects equivalent mutants.