Abstract The recent development of new portable devices enables the establishment of NIRS technology at industrial setting (on-line). However, the numerous existing NIRS databases based on a particular parameter, have been constructed with laboratory instruments (off-line), which have required a considerably effort in terms of time, labor and costs. For this reason, the transfer of calibrations between devices of different characteristics is a clearly crucial step. In this study, three different standardization algorithms: Slope/Bias Correction (SBC), Piecewise Direct Standardization (PDS) and Transfer by Orthogonal Projection (TOP) were tested and evaluated for transferring olives quality databases between an off-line NIRS monochromator (FOSS NIRSystem 6500) and a portable NIRS diode-array spectrometer (CORONA 45 visNIR). The results obtained showed that the use of TOP yielded the best Standard Error of Prediction (SEP) values for the fat content (1.97%) and free acidity (2.52%) parameters, while PDS for moisture content (2.24%). These results suggest that good calibration models for quality evaluation in intact olives can be obtained, based on spectral databases transferred between diverses NIRS spectrometers.