To discriminate samples from three varieties of Tunisian extra virgin olive oils, weighted and non-weighted multiblock partial least squares-discriminant analysis (MB-PLS1-DA) models were compared to PLS1-DA models using data obtained by gas chromatography (GC), or global composition through mid-infrared spectra (MIR). Models performances were determined using percentages of sensitivity, specificity and total correct classification. The choice of threshold level for the interpretation of PLS1-DA results was considered. PLS1-DA models using GC data gave better results than those using MIR data. Even with the most conservative threshold, PLS1-DA on GC data allowed very good predictions for Chemlali variety (99% correct classification), but had more difficulty to discriminate Chetoui and Oueslati samples (95% and 84% correct classification respectively). Non-weighted MB-PLS1-DA models benefiting from the synergy between the two sources of data were more discriminative than simple PLS1-DA, yielding better prediction for Chetoui and Oueslati varieties (98% and 90% correct classification respectively).