| Using infrared reflectance spectroscopy (NIR), the study evaluated drought resistance by four parameters including the electric conductivity after 24 h imbibitions, relative germination rate, relative seedling fresh weight and drought tolerant index with PEG treatment. Based on the 49 collected data of different drought tolerance in rapeseed with NIR, and using partial least square algorithm and multiple regression method, we established optimal calibration model, with a higher determination of 0.71～0.98 and lower standard error of 1～15.65. The results indicated no significant difference but a remarkable correlation between the two treatments, with the R2 of 0.72～0.89. The results suggest that NIR was feasible to identify the drought tolerance in rapeseed, which could be used in screening drought tolerance germplasm. |