A methodology for building a chromatographic test aiming at characterizing special base stationary phases was described. Instead of choosing its conditions a priori, a "full" comprehensive test based on extended running conditions was performed on a 12 column set. The conditions were carefully chosen from their ability to take into account the solvent and the pH effects. Principal component analysis (PCA) has been combined to hierarchical cluster analysis both to provide interpreted classifications and to reduce drastically the test itself by eliminating redundant information. The final reduced test can be considered optimal because the minimized set of test conditions allows to provide as much information as in the initial full test.