In software regression testing, reorder the test cases to ensure the modified code does not introduce new effects and keep function right. Traditional coverage-based test case prioritization (TCP) approaches are Total and Additional. These two approaches only use code coverage as the measure to rank the test cases. Several defect prediction-based TCP approaches integrate the defect prediction probability and code coverage with weighting coefficients. However, previous methods need to have a priori knowledge related to defects, which is difficult to obtain. Therefore, we proposed an approach named MFTCP by consider code coverage, defect prediction probability, and lines of code (LoC) with automatic weighting coefficients to rank regression test cases. The experimental results indicate the proposed method is more effective than the baselines, especially in terms of the total strategy.