Australia has over 30 Panicum spp. (panic grass) including several non-native species that cause crop and pasture loss and hepatogenous photosensitisation in livestock. It is critical to correctly identify them at the species level to facilitate the development of appropriate management strategies for efficacious control of Panicum grasses in crops, fallows and pastures. Currently, identification of Panicum spp. relies on morphological examination of the reproductive structures, but this approach is only useful for flowering specimens and requires significant taxonomic expertise. To overcome this limitation, we used multi-locus DNA barcoding for the identification of ten selected Panicum spp. found in Australia. With the exception of P. buncei, other native Australian Panicum were genetically separated at the species level and distinguished from non-native species. One nuclear (ITS) and two chloroplast regions (matK and trnL intron-trnF) were identified with varying facility for DNA barcode separation of the Panicum species. Concatenation of sequences from ITS, matK and trnL intron-trnF regions provided clear separation of eight regionally collected species, with a maximum intraspecific distance of 0.22% and minimum interspecific distance of 0.33%. Two of three non-native Panicum species exhibited a smaller genome size compared to native species evaluated, and we speculate that this may be associated with biological advantages impacting invasion of non-native Panicum species in novel locations. We conclude that multi-locus DNA barcoding, in combination with traditional taxonomic identification, provides an accurate and cost-effective adjunctive tool for further distinguishing Panicum spp. at the species level.