A combination of Raman spectroscopy and multivariate modeling was applied to build classification models to assess water content in soybean biodiesel. The Raman signature region of water and the characteristic vibrational modes from –CH 2 and -CH 3 moieties were used to build the models, as they are chemically influenced by water presence in biodiesel. Also, through data mining of the biodiesel digital signature region, structural modifications associated with -CH 2, CH 3, and -C=C groups were found to be related to water presence. The fluorescence influence in modeling was also studied, a common Raman interference whose origin in biodiesel can be associated with species such as vitamin E, carotenoids, and chlorophylls. Overall, an accuracy higher than 80% was reached for all models, and an improvement in the figures of merit was observed when spectra with a fluorescence background were included in the models. Furthermore, those classifications based on discriminant analysis (Partial Least Squares and Interval Partial Least Squares) reached area under the curve (AUC) values of 0.92 and 0.94, indicating a very good performance for biodiesel classification in compliance with the Brazilian legislation (ANP n° 920/2023), whose limit value of water content is currently set in 200 mg kg -1.