To improve the efficiency and accuracy of calibration in frequency domain for a distributed electromagnetic receiver, a calibration method based on multi-frequency pseudo-random signal was proposed and the generation and detection of the signals were investigated. According to the analytical expression of multi-frequency pseudo-random signals, the amplitude and distribution characteristics of the main frequency were analyzed,and a high-precision bipolar calibration source was produced through a Field Programming Gate Array(FPGA) to achieve the encoding synthesis of pseudo-random signals. Based on the noise suppression principle of correlation detection, an iterative method with multiple times for the amplitude and phase of a noisy multi-frequency calibration output signal was given,then it was validated by a simulation. Finally, the acquisition channels of the distributed dual-channel electromagnetic receiving system and a magnetic field sensor were calibrated actually. The results indicate that the amplitude error and the phase error are less than 2% and 1°,respectively, while the outside noise is not greater than the calibration output signal. The calibration method is fast, accurate and can be used in the frequency-domain calibration of distributed electromagnetic receiving systems in various field environments.