Abstract: Because the relationship between seismic attributes and reservoir characteristics is complex, the reservoir prediction using single seismic attribute often has multi-solutions. With the research going on, the classification of seismic attributes is more detailedly divided, making the optimum combination and synthetic analysis of seismic multi-attribute become necessary. Therefore, the technical workflows of multi-attribute synthetic analysis are summarized and applied to the actual data. The main steps are ① extracting and analyzing different types of attributes, ② preprocessing the extracted attributes, rejecting attribute anomalous values and filtering, ③ seeking and selecting the sensitive attributes combined with drilling and oil extraction data, ④ descending the optimal seismic attributes through K-L changes, and using the synthetic analysis technology with linearity and nonlinearity on reservoir prediction. In the process of applying this technology in one reservoir in the CN work area to forecast the oiliness, the results of this research and the results of acoustic impedance inversion have been reciprocally verified, showing that the multi-solution problem is well resolved.