Abstract : Quality of osmotically dehydrated pineapples is very important to consumers. A reliable non-destructive method for assessing quality of osmotically dehydrated pineapples is required for factories in order to ensure the quality of products be-fore sending to consumers. Near infrared hyperspectral imaging (NIR-HSI) has the potential to be used for this purpose and was therefore tested for predicting qualities including hardness, total soluble solids (TSS), sulfur dioxide content (SO2) and moisture content (MC) of osmotically dehydrated pineapples. Spectra of samples were acquired in the range of 935-1720 nm and were used for estab-lishing calibration models for the quality indices by partial least squares regres-sion (PLSR). The chemometrics was investigated in order to acquire the best models. The prediction of the models for hardness, TSS, SO2 and MC had a cor-relation coefficient of prediction (Rp) of 0.86, 0.80, 0.76 and 0.84 respectively and root mean squares error of prediction (RMSEP) of 0.82 N, 1.86%, 25.53 mg/kg and 0.73% respectively. These models were also applied to samples to ob-tain predictive images that showed spatial distributions of samples quality on a color-based scale. Results showed that NIR-HSI can be applied to use as a non-destructive quality prediction method in the on-line system of factories for sorting and guaranteeing the quality of osmotically dehydrated pineapples.