Abstract: Propagation of seismic waves through oil/gas bearing reservoirs will be affected by wave-induced flow, resulting in amplitude attenuation and elastic characteristic dispersion, and making it possible to predict the presence of fluids. However, existing theories are unclear and mathematical expressions lack accuracy, making it difficult to predict the presence of oil and gas. This study examines the attenuation of seismic amplitude and the dispersion of elastic parameters, and uses Chapman's pore-crack attenuation theory, while considering the squirt flow effect, to construct a solid-liquid decoupling fluid factor. Subsequently, the reflection coefficient characteristic equation is constructed by using the new fluid factor, and compared with the Zoeppritz and Aki approximation equations to demonstrate the improved accuracy of the new equation. Finally, a reservoir hydrocarbon-prediction method based on a pre-stack seismic inversion method using solid-liquid decoupling fluid factor is proposed. The reservoir fluid prediction is tested by using the inversion results of the new fluid factor to carry out reservoir oil/gas prediction in the well area A in the down-dropped block of Shengbei fault in China. The results show that the frequency-dependent viscoelastic solid-liquid decoupling fluid factor based on pre-stack seismic data is accurate and reliable, and the identified reservoir fluid distribution results are in good agreement with logging interpretation results. This study provides novel ideas and methods for fluid identification in complex reservoirs.