Rosewood is a kind of precious wood which include many wood species. It’s difficult to most people to identify rosewood species. Near infrared spectroscopy (NIR) of eight rosewood species was investigated in the present paper. The results showed that (1) there was significant correlation between near infrared spectroscopy and color parameters expressed by L*, a* and b* values of rosewood, the correlation coefficients between NIR predicted and laboratory measured L*, a* and b* values were 0.988, 0.991 and 0.993, respectively; (2) The eight rosewood samples can be distinctly divided into eight categories by principal component analysis (PCA), the differences in the NIR among the eight rosewood species were more distinctly revealed by the three-dimensional PCA score plot than that of the two-dimensional. The results illustrated that it was feasible to identify rosewood species by near infrared spectroscopy coupled with chemometrics, and also provided a new method to rapidly identify or classify rosewood.