In this study, independent components analysis (ICA) is tested as a mathematical method to extract characteristic signals from a data set of mid-infrared spectra acquired during accelerated oxidation of vegetable oils (corn oil, sunflower oil and soya oil) at different temperatures. These characteristic signals highlight the modifications that occur in the oils and facilitate their interpretation. ICA clearly showed that although the nature of the chemical reactions (such as oxidation and cis-trans isomerisation) induced by the heating are the same for all the types of analysed samples (corn oil, sunflower oil and soya oil), their kinetics are different depending on the oil composition. ICA also showed that higher temperatures reduce the induction times of the oxidation reaction and accelerate the degradation of the oils.