The food industry is continuously looking for reliable methods useful to standardize different control parameters and has a direct interest into bittera tasting substances, either for the identification of negative offa flavors or for the monitoring of a desired organoleptic quality. The exploitation of dedicated panel tests for sensory purposes is useful, but it suffers from limitations related to subjectivity, reproducibility, and number of analysis per day (Profile Attribute Analysis). On the contrary, sophisticated analytical solutions, such as HPLC, need trained personnel and are often too expensive or time consuming. The target of the present research work is the development of alternative techniques potentially able to detect taste molecular markers in bakery commodities, with particular attention to polyphenol detection. In particular, two different analytical approaches were developed and compared to a reference LC-MS protocol in order to detect the polyphenol concentration inside real food matrices like biscuits: FT-NIR and electroanalytical methods