Near infrared spectroscop is applied to establish a rapid detection method of osthole and columbianadin in Angelicae pubescentis Radix. Near infrared spectroscopy (NIR) has become a new analytical technique for quantitative analysis of Chinese medicinal materials because of its rapid analysis speed and no need for sample pretreatment. 97 batches of samples are collected from different regions with the determination of samples with high performance liquid chromatography (HPLC) and NIR. Due to subtle difference between the spectra, it needs to preprocess by comparing the results of different pretreatment methods to select the optimal model parameters. The results showed that the pretreatment method for the first derivative, osthole in performance optimization model, the determination coefficients was 0.941 3 while the root mean square prediction (RMSEP) was 0.141; determination coefficients of calibration set was 0.923 3, and the root mean square error of calibration (RMSEC) was 0.163. Without pretreatment of columbianadin model in optimal performance, the determination coefficients was 0.857 4 and the root mean square prediction was 0.103; when the determination coefficients of calibration set was 0.831 5, and the root mean square error of calibration was 0.112. Studies have shown that the near infrared spectroscopy combined with partial least squares method is simple, rapid, efficient and nondestructive which can achieve the goal in terms of quantitative analysis for quality control of medicines research with osthole and columbianadin in Angelicae Pubescentis Radix. At the same time a new method for the development of near infrared spectroscopy technology of traditional real-time analysis is proposed, it also has great significance to guarantee the stability of Chinese medicinal materials quality control.