Abstarct: Determination and separation of reservoir fluids is one of the critical steps in hydrocarbon reservoirs characterization. Well tests are the best means to this purpose, however their results are not easily accessible; hence, well logs are widely used recently. The main purpose of this study is to recognize gas from the other reservoir fluids (water and oil) using the sonic log and also to evaluate the effect of gas on seismic wave’s amplitude in one the oil fields in the NW of Persian Gulf. To achieve the goals, the shear wave velocity data are required at the first step that are available only for one well. Thus, different methods of linear and non-linear regression, linear and non-linear genetic algorithm have been used to estimate shear wave velocity values for other wells. The obtained results show that the value of mean square error are 0.035764, 0.0335134, 0.108179, 0.0375579 and their corresponding regression coefficients are 0.73, 0.74, 0.72 and 0.74 respectively for each of the above methods. acoording to mean square error values gained for prediction of shear and compressive wave velocity and correlation coefficient, nonlinear regression would be utilized for shear wave velocity prediction. A variety of graphs baxsed on the shear and compressive wave velocity and density have been verified to separate gas from other fluids and it has finally been found that the extended elastic impedance method is the best for this purpose. Then, in order to set a proper rock physics model for fluid substitution, some of the most common rock physics models such as: Gassmann, Hertz-Mindellin, Hashin-Shtrikman and experimental equation of Kreif are applied and compared. Finally, Gassmann model was selected as one of the most suitable models for fluid substitution in this sandstone reservoir. Next, fluid substituation is done for different value of gas saturation and their corresponding compressional velocity, shear velocity and density were determined. The results show that with presence of gas, compressional wave velocity and density of reservoir are decreased while the value of shear wave velocity is increased. Finally, seismic forward modelling has been done by using of the convolution method to determine the seismic response on seismic wave’s amplitude. The obtained results indicate that any slight presence of gas can cause considerable reduction of wave amplitude whereas the rate of amplitude decay is reduced by any increase of gas saturation so that from gas saturation upper than 15% , the wave’s amplitude is enhanced.