The selection of an appropriate solvent system is the most crucial step in high-speed countercurrent chromatography (HSCCC) separation. The compound polarity plays an important role in HPLC analysis and HSCCC separation, and it can be calculated by the HPLC polarity parameter model and the average polarity of the HSCCC solvent system, respectively. However, flow rates, columns and methanol concentrations of the HPLC experiment can influence the calculation of the compound polarity. Therefore, the applicability and accuracy of the HPLC polarity parameter model still needed to be extensively validated. We chose 14 compounds to conduct the shake-flask experiments and HPLC analysis on, such as apigenin, honokiol, phloridzin and dihydromyricetin. The HPLC analysis results showed that different flow rates and columns have negligible effects on the calculated compound polarities. However, there was a certain variation trend in the calculated polarities with different methanol concentrations. Although the polarity values of some compounds showed a difference between the HPLC analysis and shake-flask experiments, their partition coefficients (K) in the HSCCC solvent systems were still located in the range of 0.5 < K < 2.0. Guided by the HPLC polarity parameter model, the appropriate HSCCC solvent systems for mangosteen peel and Hypericum sampsonii Hance were selected, and the two main components (mangostin and quercetin) were isolated from their extracts, respectively. The separation results showed that the predicted compound polarities were sufficient to meet the HSCCC separation requirements. Meanwhile, this method required only 1 to 2 HPLC analyses with reference compounds, greatly improved the efficiency of the HSCCC solvent system selection, and shortened the experimental time. The polarity parameter model was a fast and efficient analysis method for the selection of an appropriate HSCCC solvent system.