Two statistical mixture designs were used to optimize the proportions of solvents used in both the extraction medium and the reversed liquid chromatographic mobile phase to improve the quality of chromatographic fingerprints of Bauhinia variegata L extracts. For modeling, the number of peaks was used as a measure of fingerprint information. Three mobile phases, each with a chromatographic strength of two, gave good results. A methanol/water (77:23 v/v) mixture resulted in 17 peaks in the chromatographic fingerprint whereas acetonitrile/water (64.5:35.5 v/v) and methanol/acetonitrile/water (35:35:30 v/v/v) mixtures resulted in 18 and 20 peaks, respectively. The corresponding optimum solvent compositions to extract chemical substances for these three mobile phases were ethanol/acetone (25:75 v/v/v) and dichloromethane/acetone (70:30 v/v) mixtures, and pure dichloromethane, respectively. The mixture designs are useful for understanding the influence of different solvents on the strengths of the extraction medium and the mobile phase.