Leafmining flies (Diptera: Agromyzidae) can be serious economic pests of horticultural crops. Some genera such as Liriomyza are particularly problematic with numerous species, some of which are highly polyphagous (wide host range), which can only be confidently identified morphologically from adult males. In our study, DNA barcoding was employed to establish new locality records of the vegetable leafminer fly, Liriomyza sativae, from the islands of Torres Strait (Queensland, Australia) and the central highlands of Papua New Guinea (PNG). These records represent significant range extensions of this highly invasive plant pest. Specimens of immature leafminers (from leaf mines) were collected over a 5-year period during routine plant health surveys in ethanol or on FTA® filter paper cards, both methods proved effective at preserving and transporting insect DNA under tropical conditions, with FTA cards possessing some additional logistical benefits. Specimens were identified through sequencing two sections of the cytochrome oxidase I gene and the utility of each was assessed for the identification of species and intra-specific genetic lineages. Our study indicates that multiple haplotypes of L. sativae occur in PNG, while a different haplotype is present in the Torres Strait, with genetic regionalization between these areas apart from a single possible instance – one haplotype ‘S.7’ appears to be common between these two regions – interestingly this has also been the most common haplotype detected in previous studies of invasive L. sativae populations. The DNA barcoding methods employed here not only identified multiple introductions of L. sativae, but also appear generally applicable to the identification of other agromyzid leafminers (Phytomyzinae and Agromyzinae) and should decrease the likelihood of potentially co-amplifying internal hymenopteran parasitoids. Currently, L. sativae is still not recorded from the Australian mainland; however, further sampling of leafminer flies from Northern Australia and surrounding areas is required, as surveillance for possible Liriomyza incursions, as well as to characterize endemic species with which Liriomyza species might be confused.