Abstract: 3CL proteinase plays a critical role in the replication process of coronavirus and is an important drug target. The interaction of 38 triarylpyridinones with 3CL hydrolase was studied by molecular docking method. On this basis, comparative molecular force analysis (CoMFA) was applied to study the three-dimensional quantitative structure-activity relationships of these compounds. In the CoMFA model based on the training set of 30 compounds, the cross-validation coefficient q² was 0.810, the non-cross-validation correlation coefficient r² was 0.981, and the standard deviation SEE was 0.099. The prediction coefficient r²_pred of the test set composed of 8 compounds was 0.855, which indicated that the model had good fitting ability and prediction ability. Based on the CoMFA contour maps, the favorable structural characteristics of these compounds in steric field and electrostatic field were discussed. On this basis, a group of triarylpyridinone 3CL proteinase inhibitors with good predictive activity were designed, which provides theoretical guidance for further structural optimization of these compounds.