Abstract: OBJECTIVE To quickly and accurately identify Moslae Herba and its adulterants, so as to ensure the quality of Moslae Herba and the safety of medication. METHODS The matK and ITS2 sequences of Mosla chinensis, Mosla chinensis 'Jiangxiangru' and Elsholtzia ciliate were amplified and sequenced. The sequencing results were checked by Codon Code Aligner software. The matK and ITS sequences of Moslae Herba, Mosla chinensis 'Jiangxiangru' and its adulterants Elsholtzua splendens, Elsholtzia ciliata, Elsholtzia densa, Origanum vulgare were downloaded from GenBank. The hidden Markov model was used to remove 5.8S and 28S from ITS sequences. A total of 50 ITS2 sequences of 16 species were obtained. A total of 28 matK sequences of 9 species were obtained by proofreading with Glustal software. Mega7.0 software was used for matK and ITS2 sequences analysis, intraspecific and interspecific genetic distances caculation, and neighbor joining(NJ) phylogenetic tree construction. ITS2 secondary structure was predicted by ITS2 database and aligned by 4Sale software. Profile neighbor-joining(PNJ) phylogenetic tree was constructed based on combined ITS2 sequence and its secondary structure by ProfDistS software. RESULTS The genetic distance based on matK and ITS2 sequences showed that there was an obvious barcoding gap between the authentic Moslae Herba and its adulterants. The NJ and PNJ evolutionary trees had the same topological relationship, which could effectively identify Moslae Herba and its adulterants. The ITS2 secondary structures of Moslae Herba were significantly different from its adulterants. CONCLUSION It is suggested that matK and ITS2 can be used as DNA barcodes to identify Moslae Herba and its adulterants. The addition of ITS2 secondary structure information can enrich the identification results, and provide reference for the accurate identification of Moslae Herba and the species classification of gnus Elsholtzia and genus Mosla.