Objective Optically pumped magnetometer (OPM)-magnetoencephalography (MEG) is a rapidly developing new-generation brain function imaging technology compared to the traditional MEG. Due to the proximity of the detectors and the scalp, OPM-MEG offers higher signal intensity and, with its multiaxis detection capability, can detect signals in the “blind area” of traditional MEG. It provides a powerful tool to the research of brain function and clinical diagnosis. This paper aims to investigate the signal distribution differences of biaxial OPM-MEG when measuring true physiological responses and to compare its performance with traditional MEG.Methods In this study, ten healthy subjects were examined using a 9-channel biaxial OPM-MEG during an auditory task involving frequency following responses (FFR). FFR-related magnetic responses were acquired along both tangential (Y-axis) and radial (Z-axis) directions. We analyzed the OPM-MEG data features in different axial directions and different regions, including signal intensity and regional mean energy. Additionally, we compared the results with the data detected by traditional MEG.Results After processing 800 trials, the average signal energy in the Y-axis and Z-axis was found to be 0.971 0 and 0.767 3 respectively, with no statistical significance (P=0.438). However, a regional analysis revealed distinct signal distribution patterns in the left temporal area compared to the other two regions, which was statistically significant (P=0.049). Topographical mapping showed a clear left-sided lateralization, similar to SQUID-MEG results. PSD analysis and wavelet time-frequency analysis further supported these findings. In-depth analysis of peak-to-peak values from all sensors during the stimulus period (0-0.2 s) revealed that within-participant measurements showed consistency, whereas between-participant variability was considerable. This highlights the importance of considering individual differences in future studies, as they can significantly impact measurement outcomes. The analysis revealed that the biaxial OPM-MEG signals were stronger than traditional MEG signals. Furthermore, there were significant differences in signal distribution and intensity between the two axes. In most subjects, tangential signals were found to be significantly stronger than radial signals, which are generally difficult to capture using traditional MEG.Conclusion Our findings demonstrate the capability of biaxial OPM-MEG in capturing real physiological signals and show that it provides richer information compared to uniaxial measurement. This study suggests that traditional MEG may be missing critical brain activities in its “blind areas”, highlighting the need for optimizations in brain electrical activity models based on uniaxial (radial) MEG recordings. With its multi-axis recording capability, OPM-MEG holds great potential in brain science research and the diagnosis of neurological diseases, offering a more comprehensive and precise tool for understanding brain functions.