Non-destructive freshness assessment of shell eggs during 16 days of storage at 20 degrees C was carried out by means of an FT-NIR spectrometer and a fiber optic probe. Diffuse reflectance spectra were acquired in the spectral range 833-2500 nm on samples of eggs collected from three different farms. characterized by different rearing methods. After each spectral acquisition, the freshness parameters such as air cell height, thick albumen heights and Haugh unit were destructively measured. For each rearing method, PCA (principal component analysis), PLS (partial least square regression) and PLS-DA (partial least square discriminant analysis), were carried out in order to set up models to predict the freshness parameters and to classify egg samples according to the days of storage. Hierarchical cluster analysis was also conducted to test similarity between values at different days of storage and rearing methods. Predictive models showed R(2) value up to 0.722, 0.789 and 0.676 for air cell height, thick albumen heights and Haugh unit, respectively (test set validations). Egg samples were correctly classified (100%) according to the days of storage. Hierarchical cluster analysis showed a low level of heterogeneity for the three rearing methods within the days of storage. (c) 2008 Elsevier Ltd. All rights reserved.