Surface-enhanced Raman spectroscopy (SERS) along with density functional theory (DFT) was used to develop a rapid detection method for thiabendazole residues in fresh tea leaves. Fe3O4 nanoparticles and graphitic carbon were used to eliminate the interference from fluorescent materials in fresh tea leaves, such as chlorophyll and theanine. The SERS enhancement effects of silver colloids and gold colloids for thiabendazole standard solutions were compared; the results showed that silver colloids produced better effects. By attributing and analyzing the Raman spectra of a thiabendazole molecule combined with the results of density functional theory, Raman peaks were obtained at 782, 1007, and 1576 cm-1, which could be the characteristic peaks of thiabendazole molecules. The minimum detection concentration of thiabendazole in fresh tea leaves was obtained as 1 mg/L by using the SERS method. The standard curve was plotted with the peak intensity at 782 cm-1. A good linear relationship was found between the peak intensity at 782 cm-1 and the concentration of thiabendazole within the range 0.5–20 mg/L. The recovery rate of SERS was 87.33%~93.04% and the relative standard deviation (RSD) was between 3.28 and 5.64, indicating that this method is highly accurate and reliable.