There is a pressing need to discover novel anti-tuberculosis agents to combat emerging drug-resistant strains. , and have been identified as potential sources of anti-tuberculosis (TB) drug candidates. Previous studies have examined the metabolites and metabolic pathways in mycobacterial strains affected by methanolic extracts of these plants, but the specific active compounds responsible for the antimycobacterial activity, the effect on affected metabolites and metabolic pathways of mycobacterial cell cultures, remain unclear. Untargeted metabolic fingerprinting may help identify the active compounds. The objective of this study was to model the antimycobacterial activity of methanolic extracts of , and as a function of their UHPLC-MS fingerprints and determine whether specific peaks (compounds) in the fingerprints contributed significantly to the activity. In this study, fingerprints of 18 methanolic extracts from roots, leaves, and stem barks were obtained with ultra-high-performance liquid chromatography-mass spectrometry (UHPLC-MS). The minimal inhibitory concentrations (MICs) of these extracts against a pan-sensitive strain were determined using a resazurin-based microdilution assay. Fingerprints were processed and analyzed using regions of interest-multivariate curve resolution (ROIMCR). Partial least squares (PLS) regression was employed to model the MICs. Potential active compounds, including cryptolepine (from ), verbascoside (from ), and isofagaridine (from ), were identified as antimycobacterial compounds. These compounds likely influence mycobacterial metabolic processes, including cell wall synthesis, protein production, nucleotide metabolism, and energy generation. Further investigations are required to validate our findings.