ABSTRACT Species represent an important unit for the study of diversity, but may not always be delimited consistently across regions and clades. Many of these taxonomic inconsistencies are due to the variable views of taxonomists. In recent years, however, new methodologies have attempted to circumvent this problem by assigning more objective criteria for the delimitation of species, drawing on a wide range of data such as DNA, morphology, vocalisation and ecology. Here, we apply a genetic screening approach in which we sequence the mitochondrial gene ND2 for all recognised subspecies of 16 species in eight genera (a mix of lowland and montane species) from the geologically complex tropical island of New Guinea. We show that populations within some species are genetically highly divergent despite little morphological differentiation, but we also find an example in which populations from five morphologically distinct species are genetically very similar. Overall, our data show higher levels of genetic differentiation than expected, but also highlight the difficulty of predicting which groups contain unrecognised diversity. These results are interesting in their own right, but also have implications for further analyses that focus on increasing our understanding of how diversity builds up over time.