Various studies and applied processing by businesses have been done to improve the quality of Patchouli oil (PO), such as improving appearance, reducing heavy metal content, reducing acid numbers, and increasing the ratio of Patchouli alcohol (one of the active components of PO). However, this disregards the possibility of chemical composition change which will alter the original character of the PO. This study aims to identify the slightest shift in chemical composition from adding zeolite adsorbent into the PO. The classification model was built using Fourier transform infrared (FT-IR) spectra combined with chemometrics. The used FT-IR spectra for observation are 4000-500 cm-1 using Principal Component Analysis (PCA) and Partial Least Square-Discriminant Analysis (PLS-DA). The PO samples were from five points in Gayo Lues Regency, Province of Aceh, Indonesia, with PO added with zeolite at a ratio of 0.5%, 1.0%, 1.5%, 2.0%, and 2.5% (w/v) respectively. The classification model used in this study was able to classify between PO and PO added with zeolite at each level of the ratio. Prediction with deviation and Inlier vs. Hotelling’s T2 tests provide definitive information, with the results shown by the confusion matrix.