The antitubercular activity of selected flavonoids and their structure–activity relationships were studied against Mycobacterium tuberculosis H37Rv strain radiometrically by BACTEC 460 assay. Present study led to the identification of five flavonoids, viz., luteolin, baicalein, quercetin, myricetin and hispidulin with MIC 25–100 lg ml-1 , as new antitubercular templates. Rest flavonoids were found inactive against M. tuberculosis at a concentration of 100 lg ml-1 . A possible structure–activity relationship (SAR) was also drawn to determine the specific structural requirements of flavonoids toward antitubercular activity. The hydroxyl substitution at position 5 and 7 provides no activity, whereas the hydroxyl substitutions at 5, 6, 7 (trihydroxy) or 30 , 40 (dihydroxy) are of particular importance for antitubercular activity of a flavonoid. The O-methylation or glycosylation at any of di- or trihydroxyl substitutions inactivates the antitubercular potential of the flavonoids. We have also predicted the activity of studied flavonoids through QSAR model. A multiple linear regression QSAR mathematical model was developed for activity prediction that successfully and accurately (noting the corresponding experimental activities) predicted the antituberculosis activities of studied flavonoid compounds that had the basic pharmacophore, namely luteolin, baicalein, quercetin, myricetin, and hispidulin, with experimental and predicted n log MIC (lg ml-1 ) of 3.2189 & 2.583, 3.912 & 2.433, 3.912 & 2.433, 3.912 & 3.529, and 4.6052 & 2.703, respectively. The structure–activity relationship denoted by the QSAR model yielded a very high activity-descriptor relationship accuracy of 87 % referred by regression coefficient (r 2 = 0.870533) and a high activity prediction accuracy of 81 % (rCV2 = 0.81423). These compounds may represent novel leads toward the development of pharmacologically acceptable antitubercular agent/agents.