Fourier transform infrared spectroscopy is a powerful analytical technique to study organic materials. However, in Cultural Heritage, since the sample under analysis is always a complicated matrix of several materials, data analysis performed through peak-by-peak comparisons of sample spectra with those of standard compounds is a tedious method that does not always provide good results. To overcome this problem, a chemometric model based on principal component analysis was developed to classify and identify organic binding media in artworks. The model allows the differentiation of five families of binders: drying oils, waxes, proteins, gums, and resins, taking into account the absorption bands in two characteristic spectral windows: C-H stretching and carbonyl band. This new methodology was applied in the characterization of binders in three kinds of artworks: papers of historical, archeological, and artistic value, easel paintings, and polychromed stone-based sculptures.