The Great Barrier Reef is the largest coral reef ecosystem on the planet and supports a diverse community of marine fishes, as well as the organisms that parasitise them. Although the digenetic trematodes that parasitise fishes of the Great Barrier Reef have been studied for over a century, the species richness and diversity of many trematode lineages is yet to be explored. Trigonocephalotrema, gen. nov. is proposed to accommodate three new species, Trigonocephalotrema euclidi, sp. nov., T. hipparchi, sp. nov. and T. sohcahtoa, sp. nov., parasitic in fishes of Naso Lacepède and Zebrasoma Swainson (Acanthuridae) in the tropical Pacific. Species of Trigonocephalotrema are characterised with morphological and molecular data (18S rRNA, ITS2 and 28S rRNA). Species of Trigonocephalotrema are morphologically distinguished from all other haplosplanchnid lineages by having terminal, triangular, plate-like oral suckers. With the inclusion of the new molecular data, Bayesian inference and maximum likelihood analyses of the Haplosplanchnidae Poche, 1926 recovered identical tree topologies and demonstrated Trigonocephalotrema as a well-supported monophyletic group. Although species of Trigonocephalotrema are differentiated from all other haplosplanchnid lineages on the basis of morphology, species within the genus are morphologically cryptic; thus, accurate species identification will require inclusion of host and molecular data. Species of Trigonocephalotrema cannot be assigned to a recognised subfamily within the Haplosplanchnidae using either morphological or molecular data and would require the erection of a new subfamily to accommodate them. However, we find little value in the use of subfamilies within the Haplosplanchnidae, given that there are so few taxa in the family, and herein propose that their use be avoided.