In this paper, the authors predict and describe the distribution of reservoirs and fluid in QHD29- 2 area by using pre-stack elastic impedance inversion technique based on Russell approximation. By using this method, the fluid factor f and shear modulus μ which is sensitive to fluid type can be directly extracted from elastic wave impedance inversion data volume, which connects the elastic impedance and fluid factor. The result obtained by the above method compromises high anti-noise performance of conventional elastic impedance inversion and the intuition by utilizing fluid factor for fluid identification, which effectively reduces the error cumulative process caused by the indirect computation of conventional method. The method was utilized to detect the fluid in QHD29-2 Block and achieved deeper recognition on the lateral variation of physical properties of the reservoirs.