It is essential to understand how non-native species have invaded and modified native community. Recent advances in metabarcoding technique enable us to examine structure of biological community more efficiently. However, because DNA database for species identification is incomplete, especially in highly diverse arthropod species, native and invasive status remain unknown in many species. In this study, we developed a novel approach to distinguish invasive species from native species by using genetic information from the metabarcoding dataset. We collected arthropod samples (insects, spiders, isopods etc.) in the Hawaii island and identified the taxonomic name of some species by blast search. We showed that intraspecific genetic variation was lower and genetic distance to the neighbor species was longer in invasive species than in native species. Moreover, we detected differences in body size and distribution range between invasive and native species. Based on this information, we predicted the invasive or native status of the unidentified species by a generalized linear model.