Abstract: In the wind tunnel sound source localization test, airflow interference makes it difficult for the array to measure at close range. Due to the limitation of the Rayleigh criterion, the traditional beamforming method is difficult to achieve high-resolution localization for multiple low-frequency sound sources with close distribution in the far field. To solve this problem, an iterative weighted off-grid localization method based on equivalent flow velocity correction is proposed. First, an off-grid model is established, with the sound source intensity and position offset compensation as the variables to be solved. Then, the block transfer function in the flow field environment is corrected by the equivalent flow velocity method. Finally, based on the iterative weighted method as a sparse constraint, the source intensity and off-grid deviation in the block sparse solution vector are accurately estimated, and the off-grid sound source is accurately identified. The method is verified in the 8 m × 6 m acoustic wind tunnel at the China Aerodynamics Research and Development Center, where an aircraft model is equipped with speakers for sound source identification experiments. The results show that in the range of 40 m/s flow velocity, the sound source position error of the traditional method is more than 0.27 m, and the amplitude error is more than 5 dB, while the position and amplitude errors of the proposed method are reduced to 0.17 m and 3.9 dB, respectively.