Characterizing drought resistance of forest trees is an important aim in the worldwide context of increasing forest decline and dieback. Furthermore, wood demand is increasing all around the world, urging the need for the domestication of native species and the selection of the most resistant individuals. While near-infrared spectroscopy is now recognized as an efficient tool to predict wood physical and chemical properties involves in wood quality, its ability to predict wood hydraulic function remains anecdotic, although it has recently gained interest. In this study, we tested near-infrared spectroscopy as a high-throughput method for quantifying both resistance to cavitation and basic wood density in the South American conifer Austrocedrus chilensis. Models were obtained in two laboratories using the same sample set. To assess resistance to cavitation, vulnerability curves were obtained using the air-injection method. Basic wood density was measured in the same samples as used for resistance to cavitation. Partial least squares regression models with cross-validation were used to establish relationships between the NIRS spectra and wood traits. The accuracy of the models strongly depended on the traits being used for calibration. Models obtained in both laboratories were similar, demonstrating the genericity of the approach whatever equipment used. Promising results were obtained for two cavitation parameters (water potential at 12% and 50% of hydraulic conductivity loss) and basic wood density. Prediction errors for these traits were low. In contrast to studies at the interspecific level, basic wood density and resistance to cavitation were negatively correlated at the intraspecific level in the studied species, but our calibration models shown than basic wood density cannot be used as a good predictor of functional traits. The NIRS has high potential to be put into practice as a rapid, reliable, and non-destructive method to determine, resistance to xylem cavitation, a trait involved in drought resistance.