Biodiversity studies are often limited by unavailability or inaccessibility of taxonomic expertise; in New Zealand, taxonomic revisions and keys to many invertebrate groups are far from complete. To make progress with ecological and biodiversity studies, the separation of organisms into recognisable taxonomic units or morphospecies has sometimes been adopted. Coleoptera are speciose, trophically diverse, and taxonomically well known compared with other large trophically diverse groups and so they are useful to include in biodiversity studies. This study opportunistically examined the accuracy of Coleoptera species separation using morphologically recognisable features of specimens collected from three different vegetation communities, by three student researchers with different levels of training and previous expertise. Their morphospecies were examined by a single researcher with experience in taxonomy of Coleoptera. In total, 155 morphospecies were separated by the three students, compared with 151 determined by the specialist, which included representatives from 23 families of Coleoptera. All three students identified a total number of morphospecies within about 10% of the actual number, irrespective of previous training. However, the proportion of correct species separation increased from 63 to 87% in accordance with the level of previous experience. Common errors in species separation made by parataxonomists in relation to coleopteran families are discussed.