Herbal medicines have an important role in clinical therapy in Asian countries such as Korea, Japan, and China. The objective of this study is to develop a nondestructive and accurate analytical method to discriminate herbal medicines according to geographical origin. Even though they are the same species, their qualities are different by growing conditions such as climate and soil. Near infrared (NIR) reflectance spectroscopy and a pattern recognition technique were applied for discrimination of herbal medicines according to geographical origin (Korea and China). Astragali Radix (AR), Ganoderma, and Smilacis Rhizoma (SR) were examined. It is shown that the representative NIR reflectance spectra in each group are different according to geographical origin after second derivatization to enhance spectral features. Also, the NIR reflectance spectra of Chinese and Korean samples were differentiated using principal component (PC) score plots. To establish the discrimination rule, Mahalanobis distance and discriminant analysis with PLS2 were utilized.