Current technological advances are resulting in an increased interest in computational methods for predicting the physicochemical properties of biological active compounds, such as lipophilicity, for example. There is a strong need to develop new and accurate in silico models that can be based on structural descriptors such as topological indices, which are sets of numerical descriptors that describe the molecule under study. The arrangement of atoms in a molecule is closely related to its topology and geometry, which correlates with the pharmacokinetic properties of the substance, such as ADME/T. In this study, Wiener (W), Randić (⁰χ, ¹χ, ⁰χ^ν, ¹χ^ν), Gutman (M, M^ν), Pyka (A, ⁰B, ¹B) and Rouvray-Crafford (R) topological indices were calculated for selected antimicrobial compounds such as delafloxacin, linezolid, sutezolid, ceftazidime and selected immunosuppressive compounds like everolimus and zotarolimus. Linear regression analysis was used to create linear correlations between the calculated topological indices and the values of lipophilicity parameters previously obtained by TLC technique and calculated by computer algorithms. Our work indicates that structural descriptors like topological indices can be a useful tool for predicting selected important ADME/T properties of drugs, such as lipophilicity. The best predictive power (r>0.9) indicate the linear models based on the following topological indices: R,W,A. The proposed method is fast, easy to use, and economical because it avoids expensive laboratory experiments to study ADME/T properties by experimental methods.