Background: For the past 25 years, the ITS rDNA (ITS1-5.8S-ITS2) of Gyrodactylidae has been crucial for species identification, description, and phylogeny. This family includes 25 genera parasitizing marine and freshwater fish, initially distinguished by morphological differences in attachment and/or male copulatory organs. Gyrodactylus Nordmann, 1832, the most species-rich genus, has approximately 500 described species and an additional 25,000 species suspected. The genus is not monophyletic, and the functionally adaptive nature of morphological diagnostic characters complicates the delimitation of new genera. Methods: A phylogeny based on ITS rDNA was proposed to address these challenges, using only complete sequences of primitive taxa. Fifty-four sequences were aligned with the MUSCLE v5.1 algorithm, creating a 1590 ps long matrix. Maximum Likelihood (ML) and Bayesian Inference (BI) methods with the models TVM+F+G4 and SYM+G4 for ITS1–ITS2 and 5.8S, respectively, were inferred using IQ-TREE v2.3.5 and BEAST v2.7.6.0. Results: The findings revealed eleven main lineages. Four of them are proposed for classification into new genera: Cichlidarus gen. nov., Iraqemembranatus gen. nov., Macracanthus gen. nov., and Rysavyius gen. nov. Elevating the subgenus G. (Gyrodactylus) to genus rank was supported. Conclusions: The presented phylogeny provides a foundation for developing a classification system within Gyrodactylidae that is both reasonable and comprehensive.