Lipophilicity is considered to be the most important physicochemical property, in addition to the ionization constant and solubility, as it has a direct impact on the absorption, distribution, metabolism and excretion (ADME) properties of a potential drug in the body, as well as on toxicity. It represents the affinity of a molecule or part of a molecule for a lipophilic environment and expresses the logarithm of the partition or distribution coefficient (log P or log D). Due to the significant impact on the properties of the potential drug, it is usually determined already in the very early stages of drug discovery. Traditional methods of determining lipophilicity, such as the shake-flask method, are limited to soluble compounds and require a larger amount of compound and a longer time. Therefore, they are being replaced by indirect high-throughput analyses, such as reversed-phase high-performance liquid chromatography (HPLC) in the early stages of discovery. Such an indirect measurement is the method for determining the chromatographic hydrophobicity index (CHI), which is based on calculating CHI values from gradient retention times of the analyte using the equation obtained by linear regression analysis. The aim of this work was to develop a method for determining CHI under acidic (pH 2.7) and basic (pH 10.5) conditions of the mobile phase using reversed phase ultra high performance liquid chromatography (RP-UPLC) coupled to mass spectrometry (MS). CHI was determined at pH 2.7, 7.4 and 10.5 for a set of 30 compounds, divided into the following groups: acidic, basic, neutral, amphoteric and macrolides. In silico values for log D at pH 2.7, 7.4 and 10.5 were calculated using a commercial software ACD/Percepta. A method with high precision for determining CHI at pH 2.7 and pH 10.5 was successfully implemented using RP-UPLC/MS. The influence of an aging chromatographic column in the method was also shown. By determining of CHI values at three different pH values it was possible to characterize the acid-base character for all tested compounds. A good correlation between measured CHI values and with in silico predictions was obtained for most compounds.