Palm oil, soy oil, sunflower oil, corn oil, castor oil, and rapeseed oil were analyzed with Fourier transform infrared (FT-IR) and FT-Raman spectroscopy. The quality of different oils was evaluated and statistically classified by principal component analysis (PCA) and a partial least squares (PLS) regression model. First, a calibration set of spectra was selected from one sampling batch. The qualitative variations in spectra are discussed with a prediction of oil composition (saturated, mono- and polyunsaturated fatty acids) from mid-infrared analysis and iodine value from FT-Raman analysis, based on ratioing the intensity of bands at given wavenumbers. A more robust and convincing oil classification is obtained from two-parameter statistical models. The statistical analysis of FT-Raman spectra favorably distinguishes according to the iodine value, while the mid-infrared spectra are most sensitive to hydroxyl moieties. Second, the models are validated with a set of spectra from another sampling batch, including the same oil types as-received and after different aging times together with a hydrogenated castor oil and high-oleic sunflower oil. There is very good agreement between the model predictions and the Raman measurements, but the statistical significance is lower for mid-infrared spectra. In the future, this calibration model will be used to check vegetable oil qualities before using them in polymerization processes.