The state of the art of turrid (=Turridae s. l.) systematics is that shells — when they include the protoconch — are reliable species-level identifiers, but inadequate proxies for allocation to genera or subfamilies. Generally, the radula is used for allocation to a (sub)family, but the hypothesis that the radula is a more adequate proxy than the shell for relationships has not yet been tested by molecular data. Species of Xenuroturris may have drastically different radulae, with either “semi-enrolled” or “duplex” marginal teeth, although their shells are very similar or even almost indistinguishable. Molecular data confirm that specimens with different types of radulae constitute different species, but two species of a pair with respectively semi-enrolled and duplex teeth end up being not closely related. However, it is still unresolved whether species with semi-enrolled (=Iotyrris) and duplex teeth (=Xenuroturris) form two supported monophyletic clades. Iotyrris devoizei n.sp. and I. musivum n.sp. are described from Vanuatu, where they occur sympatrically with I. cingulifera and Xenuroturris legitima.