The potential of fluorescence spectroscopy for predicting analytical, rheological and baking parameters of twelve wheat flours were investigated. Partial least square regression models coupled with genetic algorithm were applied on spectral data to optimize the prediction of the aforementioned quality parameters using different pre-processing methodologies. Good linear regression models were obtained for protein, wet gluten and the sedimentation value from the analytical parameters group with a R-2 of 0.90, 0.92 and 0.81 respectively. Similarly prediction was obtained for rheological parameters like the dough development time and water absorption, with a very low root mean square error of cross validation (RMSECV) and an optimal R-2 of 0.95 and 0.77 respectively while it settled at 0.78 for pasting temperature. Furthermore, baking parameters like the moisture and volume of bread were predicted with a decent accuracy showing a R-2 of 0.86 and 0.95 respectively. Hence, fluorescence spectroscopy can be used as rapid method in predicting the wheat quality and its baking characteristics by just taking the spectra of flour with no sample preparation. (C) 2016 Elsevier Ltd. All rights reserved.