QSAR is one of the most popular methods that pharmaceutical industry uses nowadays to find new therapeutic agents and in their optimisation. The primary purpose of this method is to determine the correlation of chemical structure of the series of compounds and a specific chemical or biological process. The aim of this study was to show the basic approaches that pharmaceutical industry uses nowadays in the design of new drugs and to give insight into their basic features. A special emphasis was put on the QSAR approach, the topological indices and their importance in the modeling of a new drug. Within the QSAR method there are numerous approaches (e.g. molecular descriptors) that individual molecules can be characterised with. One of most important molecular descriptors in the QSAR method is the partition coefficient between octan-1-ol and water (phosphate-buffer, pH = 7,4), logP. The numerical index that characterises molecules is called topological index. It numerically expresses the topological information for the compound, which gives some indication of its size and shape, i.e. the nature of the chemical structure. To date, a whole range of topological indices that scientists use on a daily basis is known, but in this work we are dedicated to basic, well-known indices that very well describe the structure of molecules. These are: Wiener's number, Randic's index, Balaban's index, Trinajstic's number and information-theoretical index. The most modern approach to drug design is the method of molecular modeling. This method involves drug design using computer, CADD (engl. Computer-Aided Drug Design). : The obtained values of topological indices were compared to the physico-chemical properties of series of compounds from the group of sulfonamides and we noticed some connections. The values of Wiener's number, Randić's, Balaban's and information-theoretical indices are proportional to the number of peaks of molecules and to the molecular weight. The application of the QSAR method in the drug design process has greatly improved the design process itself and has enabled identification of an appropriate molecule before its synthesis.