Lamiaceae, widely used as herbal medicine, is increasingly vulnerable to adulteration driven by market demand, compromising product safety and efficacy. Prevention is challenging due to morphological similarities; thus, DNA-based phylogenetics offer an alternative for accurate species authentication. However, Lamiaceae phylogenetics remain complicated by inconsistencies between morphological and DNA data. This study reconstructed Lamiaceae phylogeny using partial ITS, matK, and rbcL barcodes to evaluate their potential application in species authentication and adulteration prevention. Sequences for 52 species across 11 genera (Spathodea campanulata: outgroup) were obtained from NCBI GenBank, aligned, and trimmed. Four maximum parsimony (MP) trees were constructed in MEGA 11 (three single-barcode, one concatenated). The concatenated dataset was also analyzed by maximum likelihood (ML). Tree robustness was evaluated with bootstrapping, consistency index (CI), and retention index (RI). matK had the longest mean sequence (785.6 bp), rbcL the highest homology (83.5%), and ITS the most parsimony-informative sites (40.3%). MP trees exhibited moderate homoplasy (mean CI = 0.63) but strong synapomorphic signal (mean RI = 0.83). Individual barcodes produced similar genus groupings, yet misplaced several species. Concatenation corrected these positions across MP and ML trees, resolving six robust monophyletic clades (bootstrap >70%), broadly consistent with earlier phylogenies: Callicarpa; Scutellaria; Clerodendrum, Lamium, and Stachys; Salvia; Thymus, Origanum, and Mentha; Orthosiphon and Ocimum. Topological discrepancies with prior studies likely reflect differences in barcode choice and taxon sampling. Concatenated barcodes improved phylogenetic resolution in Lamiaceae, producing clades that identify potential adulterants and guide DNA marker development for species authentication and adulterant detection.