Abstract: Gas content is an important index for evaluating resource potential of shale gas. Up to date, the seismic method has been adopted as the main tool for shale gas evaluation. However, it is mostly qualitative. There is so far no matured method for quantitative prediction of shale gas based on seismic data. Through the mathematical statistics of the core-measured gas content and other geophysical parameters, we selected 4 elastic parameters, namely the density, P-wave impedance, P-S wave velocity ratio and λρand established a multiple linear regression model suitable for the Jianshiba shale gas field. The density, P-wave impedance and other relevant elastic parameters are obtained through pre-stack simultaneous inversion. And quantitative prediction of gas content is made upon the results. The predicted and the measured values of gas content match well and have little errors. It is found out in this paper that the 1st Submember at the bottom of the Longmaxi Formation-Wufeng Fomation has higher gas content, and is the most favorable producing layer in the region.