A substantial portion of biodiversity has evolved through adaptive radiation. However, the effects of explosive speciation on species interactions remain poorly understood. Metazoan parasites infecting radiating host lineages could improve our knowledge because of their intimate host relationships. Yet limited molecular, phenotypic and ecological data discourage multivariate analyses of evolutionary patterns and encourage the use of discrete characters. Here, we assemble new molecular, morphological and host range data widely inferred from a species‐rich lineage of parasites ( Cichlidogyrus , Platyhelminthes: Monogenea) infecting cichlid fishes to address data scarcity. We infer a multimarker (28S/18S rDNA, ITS1, COI mtDNA) phylogeny of 58 of 137 species and characterize major lineages through synapomorphies inferred from mapping morphological characters. We predict the phylogenetic position of species without DNA data through shared character states, a morphological phylogenetic analysis, and a classification analysis with support vector machines. Based on these predictions and a cluster analysis, we assess the systematic informativeness of continuous characters, search for continuous equivalents for discrete characters, and suggest new characters for morphological traits not analysed to date. We also model the attachment/reproductive organ and host range evolution using the data for 136 of 137 described species and multivariate phylogenetic comparative methods (PCMs). We show that discrete characters not only can mask phylogenetic signals, but also are key for characterizing species groups. Regarding the attachment organ morphology, a divergent evolutionary regime for at least one lineage was detected and a limited morphological variation indicates host and environmental parameters affecting its evolution. However, moderate success in predicting phylogenetic positions, and a low systematic informativeness and high multicollinearity of morphological characters call for a revaluation of characters included in species characterizations.