Surface-enhanced Raman spectroscopy (SERS) using gold (Au) nanoparticles and NaCl solution as SERS enhanced substrate was applied to rapidly identify furaltadone metabolite (AMOZ) and nitrofurantoin metabolite (AHD) residues in chicken samples. The background interference was eliminated by using adaptive iterative re-weighted penalized least squares (air-PLS), and the spectra data was preprocessed by standard normalization. The samples of four groups, i.e. chicken extract containing AHD, AMOZ, AHD and AMOZ, and control, were marked in sequence, and the identification model was established by combining the principal component analysis and linear discriminant analysis (PCA-LDA). The experimental results showed that the correct discriminant rate of the calibration set was 90.48%, and the correct discriminant rate of the prediction set was 94.29%. Consequently, AHD and AMOZ residues in chicken samples could be identified effectively by using SERS and PCA-LDA.