On the base of absorption spectra in the range from 0.2 to 2.6 THz of seven common dangerous goods using terahertz time-domain spectroscopy (THz-TDS) technique, the THz absorption spectra of seven dangerous goods were identified successfully by fuzzy recognition. Using different feature absorption peaks of explosive and illegal drugs such as RDX, γ-HNIW, DNT, MA and Ketamine as data source of fuzzy cluster analysis, fuzzy similar matrix was built by correlation coefficient, fuzzy equivalent matrix was obtained by the method of transitive closure, standard model bank of THz absorption spectra was formed, and data were preprocessed by range analysis and compute Hamming approach degree. It was concluded that uninspected goods were RDX hidden behind the uniform and MDA. The research indicated that different feature absorption by interactions between molecules and phonon resonance mode is the basis for determining the type and category of dangerous goods, it is feasible to apply fuzzy recognition to the identification of dangerous goods, providing an effective new method for the secure inspection and identification of threats using THz-TDS technique.