A prediction model was established for the rapid analysis of carbofuran residue in duck meat by synchronous fluorescence spectroscopy method，based on the condition that the strong fluorescent derivatives can be generated in the reaction between carbofuran hydrolyzate and phthaldialdehyde（OPA） in the presence of mercaptoethanol.The 3D synchronous fluorescence spectra of the duck meat containing carbofuran were analyzed，and 120 nm was selected as the optimum wavelength difference.19 wavelengths between 240 nm and 450 nm were selected as the input features for quantitative analysis models using genetic algorithm combined with the root mean square error of cross-validation.The performances of three regression models，SVR,PCR and PLS were compared，and the results showed that the prediction of SVR regression model was the best，the determination coefficient（r2） and the root mean squared error of prediction（RMSEP） for the prediction samples were 0.999 4 and 0.878 7，respectively.The results approved that the method of synchronous fluorescence analysis with SVR had a good performance and accuracy in quick test，and was applicable for the detection of carbofuran residue in duck meat.