Abstract: In order to identify the inner and outer defect signals in pipeline magnetic flux leakage (MFL) internal inspection under dynamic conditions, a mathematical model of dynamic leakage magnetic field based on the magnetic multipole field was established for the induced eddy current magnetic field generated in the pipe wall under dynamic conditions. With the data characteristics of the inner and outer defects of pipeline extracted from the measured defect signals obtained by the inline MFL detector, the method for distinguishing the inner and outer defects of the pipeline based on the magnetic multipole field was determined: that is, the defect can be judged as an inner defect if the rising edge or the falling edge of the defect, or both, have the maximum absolute value of the second-order magnetic field parameter, or as an outer defect if both of the rising edge and the falling edge do not have the maximum absolute value of the second-order magnetic field parameter. Additionally, the method was verified by the identification probability of the measured data of the inner and outer defects. The results show that, the new distinction method has a high recognition accuracy rate, which breaks through the conditional restrictions of detector's relying on hardware sensors. It is of guiding significance to the identification of MFL defects, principle analysis and distinction of inner and outer defects.