In this study, we used a terahertz time-domain spectroscopy system to evaluate the terahertz spectra of the baicalin mixtures and observed that the mixtures exhibited obvious absorption characteristics in 0.3--1.5 THz. Further, we applied two-dimensional correlation spectroscopy to analyze the terahertz spectra of the mixtures, which were sensitive to the variation in the concentration of baicalin. Herein, two quantitative analysis models were established using the support vector machine (SVM) and partial least squares (PLS) methods, respectively, based on the aforementioned spectral analysis. Subsequently, the KS method was applied to select the calibration set and the verification set for the two models, which were evaluated using the correlation coefficient and the root mean square error (RMSE). The experimental results indicate that a good correlation between the data predicted using two quantitative analysis models based on the SVM and PLS methods and the actual data. In this case, the RMSEs are small. Baicalin can be quantitatively detected by the quantitative analysis models based on SVM and PLS. The predictive result of the SVM model is better than that of the PLS model. This study provides a new and significant method for the quality detection of medicine.