A series of 82 5-aryl-2-thio-1,3,4-oxadiazole derivatives were screened for their anti-mycobacterial activities against Mycobacterium tuberculosis H37Rv. The synthesized compounds 30-37 appeared to be the most active derivatives exhibiting more than 90% inhibition of mycobacterial growth at 12.5 μg/mL. Structure-activity relationships study was performed for the given series by using the electronic-topological method combined with neural networks (ETM-NN). A system for the anti-mycobacterial activity prediction was developed as the result of training associative neural network (ASNN) with weights calculated from projections of a compound and each pharmacophoric fragment found on the elements of the Kohonen's self-organizing maps (SOMs). From the detailed analysis of all compounds under study, the necessary requirements for a compound to possess antituberculosis activity have been formulated. The analysis has shown that any requirement's violation for a molecule implies a considerable decrease or even complete loss of its activity. Molecular docking studies of the compounds allowed shedding light on the binding mode of these novel anti-mycobacterial inhibitors.