The tight gas reserves of He8 member in Linxing area of Ordos Basin are large. However, the testing production of most wells in this member is low at the present stage. Recently, three wells have been tested, and the production yield after fracturing is more than 10000 m³/day, which reveals that the He8 member is of potential for development. According to the analysis of drilled wells, porosity is the key factor affecting productivity. Therefore, searching for high porosity reservoirs is of great significance for improving productivity of He8 Member. The rock physics analysis shows that V_p/V_s ratio in this area can distinguish sandstone from mudstone. According to the rock physics analysis result, this study predicts the distribution of sandstone in He8 member of Block A in Linxing area through pre-stack inversion. On the basis of pre-stack inversion result, the porosity prediction technology of tight sandstone reservoir is used to predict the porosity of thick reservoir by linear fitting for different reservoirs seperately for data in low P-impedance area. For the high impedance area, Bayesian discriminant analysis and co-simulation are used to predict the porosity of thin reservoirs. The porosity prediction results of thick reservoir and thin reservoir are combined to obtain the overall porosity prediction results of the reservoir. The results show that this method can effectively predict the porosity distribution of Class Ⅰ thick reservoirs and Class Ⅱ thick reservoirs, and simultaneously depict the porosity distribution of Class Ⅰ and Class Ⅱ thin reservoirs. By means of using different prediction method for different types of reservoirs, the description of reservoir porosity is more accurate, providing effective results for efficient development of tight sandstone reservoirs.