The taxonomy of trematodes of Great Barrier Reef (GBR) fishes has been studied in some detail for over 20 years. Understanding of the fauna has been informed iteratively by approaches to sampling, understanding of morphology, the advent of molecular methodology and a feed-back loop from the emergent understanding of host specificity. Here we analyse 658 host-parasite combinations for 290 trematode species, 152 genera and 28 families from GBR fishes. These are reported from 8 orders, 38 families, 117 genera and 243 species of fishes. Of the 290 species, only 4 (1·4%) have been reported from more than one order of fishes and just 23 (7·9%) infect more than one family; 77·9% of species are known from only one genus, and 60% from only one species of fish. Molecular studies have revealed several complexes of cryptic species and others are suspected; we conclude that no euryxenous host distribution should be accepted on the basis of morphology only. The occurrence of individual trematode species in potential hosts is patchy and difficult to predict reliably a priori or explain convincingly a posteriori. These observations point to the need for a vigorous iterative interaction between the accretion of host specificity data and its interpretation.