The aim of this work was to evaluate the suitability of Near-Infrared Hyperspectral Imaging Spectroscopy (NIR-HSI) for the quantification of water content in commercial biscuits. Ten biscuits from two commercial brands were conditioned for one week in desiccators with different water activities controlled by different saturated solutions of salts. Biscuits were analyzed by NIR-HSI and resulting images are analyzed by Principal Components Analysis (PCA). A regression model is also computed using a small number of variables in order to predict moisture values for each pixel and reconstruct prediction images in false color. These results obtained by PCA show that the water content of the two brands of biscuits can be easily monitored by NIR-HSI. The determination coefficient (R-2) obtained by regression is higher than 0.925 and the Root Mean Square Error of Validation (RMSEV) less than 1.815. Prediction images are very relevant and can be used to study biscuits defects.