Abstract: Magnetic flux leakage testing (MFL) is a nondestructive testing method widely used in defect detection and evaluation of steel structures. The inversion of defect signal is the key part of MFL testing, and effective defect depth profile optimization is very important to improving the efficiency of defect inversion. In this paper, an iterative optimization method of defect depth profile based on the depth characteristics of MFL signal was proposed. Based on the approximate linear relationship between the average value of MFL signal and the depth, the initial depth sequence of defect was estimated, and the defect depth sequence was optimized and updated according to the error analysis between MFL signal and MFL signal. The defect depth profile was reconstructed by inversion. Through the simulation and experiment, the MFL detection signals of different types of defects were inversed. The simulation and test results verify the effectiveness of the method.