Abstract: This research presents an Agarwood oil grading system using Support Vector Machine (SVM). Agarwood is grown in tropical parts of Asia (including Malaysia) and is a valuable international commodity. It is used primarily in fragrance and medicine. Data collected from 96 Agarwood oil samples of different qualities were used to train several SVMs with different Kernel functions. Implementation of the project was done using MATLAB v2010a. It was found that nonlinear Kernels were able to produce 100% accuracy, outperforming the linear Kernel (87.5% accuracy).