The results of petro-physical research show that pre-stack elastic parameters and multi-parameter intersection are more sensitive to reservoir and hydrocarbon,and the pre-stack inversion based on this conclusion is an effective method for quantitative prediction of the oil-bearing reservoir.As for compact and thin sand reservoir and its oil-bearing property,because their seismic response are extremely concealed and the seismic resolution is deficient,the high-resolution pre-stack stochastic inversion based on variogram is widely applied.However,the high frequency contents of pre-stack stochastic inversion have strong randomness and low reliability,which exerts a great impact on the accuracy and reliability of the prediction results.In order to solve these problems,the authors took the prediction for compact and oil-bearing thin sand of Jurassic Shaximiao Formation in north-central Sichuan Basin as a study case and,after analyzing the sensitive elastic parameters of reservoir and its oil-bearing property,carried out the application research on the high-resolution pre-stack inversion method based on seismic waveform indicator.The method uses the similarity of pre-stack seismic waveform instead of variogram to optimize well samples,and builds the initial model of each elastic parameter by stochastic simulating based on the optimization samples.Under a Bayesian framework,the method combines the initial model,the prior probability distribution of well elastic parameters and the inversion elastic parameters of pre-stack simultaneous inversion to build the posterior probability distribution of each elastic parameter,then resamples the posterior probability distribution function using Metropolis-Hastings method to get a convergent and stationary distribution which is also an inversion result.The practical application shows that it can get a high-resolution and broadband result of each elastic parameter in the formation from the pre-stack seismic data by using the pre-stack inversion based on seismic waveform indicator,and the high frequency contents of the result are more certain to raise the prediction accuracy and reliability of compact and thin sand reservoir and its oil-bearing property.It has extensive application prospect as an effective method for high-precision prediction of the reservoir and the oil-bearing property of compact clastic rocks in terrestrial facies.