Pinus pinea L is one of the most important nut species in the world given the high nutritional and culinary value of its seed, the pine nuts, with increasing demand. The objective of this study was to assess the potential of visible and near infrared spectroscopy (VIS + NIRS) to analyse different sample presentations and to discriminate geographical origins of Mediterranean pine nut grown in Chile. Pine nuts were collected from 76 adult trees in three grovith macrozones previously defined for stone pine in Chile. Original spectroscopic data were obtained by means of a Foss NIRSystems 6500 SYII spectrophotometer using a transport module. Reflectance was employed in the wavelength range of 400-2500 nm. The best means of sampling for the pine nuts used in this study were also studied. After analysing the spectroscopic data, discriminant models were obtained by means of discriminant partial least square (DPLS) with all samples. For the three macrozones previously identified, 87.8% of samples was correctly classified in the cross validation stage with the best model for pine nuts spectra analysis, obtained with shelled intact pine nuts. Results indicate that NIRS technology is capable of differentiating between pine nut samples of different geographical origins with errors ranging between 12.2 and 9.2%, and demonstrate the potential of VIS NIRS technology as a rapid and accurate method for predicting the geographical origin of Mediterranean pine nuts. (C) 2016 Elsevier Ltd. All rights reserved.