A multivariate sensor based on mid-infrared spectral signals have be used for detection of highly hazardous materials (HHMs) employing chemometrics tools. The HHMs used were the nitroaromatic compounds 2,4,6-trinitrotoluene, the aliphatic nitrate ester pentaerythritol tetranitrate and the aliphatic nitramine hexahydrotrinitrotriazine. HHMs were deposited on real-world substrates such as aluminum, cardboard, travel bags and wood. Multivariate analysis by Partial least squares (PLS) regression analysis combined with discriminant analysis (PLS-DA) was used to discriminate, classify, and identity similarities in the spectral datasets. The results show that the Multivariate vibrational detection investigated herein for multivariate sensor development is useful for the detection of HHMs on the types of real-world surface studied.