Abstract: As an important unconventional hydrocarbon resource, volcanic reservoirs hold significant potential for exploration and development. However, their complex lithological composition, tight reservoir properties, and strong heterogeneity in fluid distribution pose significant challenges for reliable fluid prediction. Based on the analysis of rock physics response mechanisms, a frequency scanning mechanism-based method for inverting the fluid bulk modulus dispersion attribute (FS-AVO) was proposed by utilizing the dispersion response characteristics of seismic waves during their propagation through subsurface media, thereby enhancing the accuracy of fluid prediction in volcanic reservoirs. A frequency-dependent reflection coefficient formula related to the fluid bulk modulus was derived. By introducing a frequency scanning mechanism, a dispersion attribute inversion process based on the FS-AVO method was constructed, achieving effective prediction of the fluid bulk modulus dispersion attribute D_Kf-max. Theoretical model tests demonstrate that compared to that calculated by the conventional dispersion inversion method (FD-AVO), the D_Kf-max calculated by the FS-AVO method exhibits higher sensitivity to changes in gas bearing within the reservoir. Application results based on actual seismic data demonstrate that this method effectively enhances the accuracy of gas-bearing prediction in volcanic reservoirs, providing a reliable basis for fluid identification in such complex volcanic reservoirs.