Ecological studies that enhance our understanding of the structure and function of the natural world rely heavily on accurate species identification. With rapid sample accumulation and declining taxonomic expertise, cladistics, phylogenetics and coalescent-based analyses have become key tools for identification or discrimination of species. These tools differ in effectiveness and interpretation depending on researcher perspective and the unique evolutionary histories of the taxa. Given the cost and time required for taxonomic assessment of ambiguous species groups, we advocate a pragmatic approach to clarify species assignment. We carried out a case-study on species from the diverse ophiuroid genusOphiacanthacommon in shelf habitats around the Southern Ocean. Although several of the species are formally described with clear and distinctive morphological characters and reproductive strategies (O. vivipara,O. pentactis,O. densispina,O. antarctica, andO. wolfarntzi), recent molecular data has highlighted issues with these morphospecies, the characters that formally define them and their evolutionary histories. Here we provide evidence that key morphological features of species can be deceptive and show that six-armedO. vivipara, for example, is not a widely distributed Southern Ocean species as currently accepted, rather, three disparate clades.Ophiacantha pentactis, described as having five arms, frequently has six arms and the six-armed form is mistakenly classified asO. vivipara. All six-armed specimens collected from the Antarctic continental shelf fall into theO. pentactisspecies clade. Molecular tools designed for species delimitation appear to fail to reflect the “true” species composition. Rather than rely on a single tool for species recognition, we advocate an integrated approach using traditional detailed taxonomic morphology, summary statistics of molecular sequence data from populations, robust phylogenies, sufficient geographical sampling and local biological knowledge to ensure that species hypotheses can be built on mutually supporting lines of evidence.