Approaches enabling prediction of chocolate quality from cocoa composition would avoid time- and money-consuming steps to chocolate makers. Average mass spectra of cocoa-polyphenol-extracts led to fingerprints used to select the molecules that discriminate chocolate sensory groups.16 worldwide cocoa samples were processed into chocolates which were characterized by sensory analysis, allowing sorting of the samples into four sensory groups.The cocoa polyphenol extracts were analyzed by liquid chromatography−low-resolution mass spectrometry. Averaging each mass spectrum provided polyphenolic fingerprints, which were combined into a matrix and processed with chemometrics (PCA, PLS-DA) to select the most meaningful molecules for discrimination ...