Although the differentiation of clades at the species level is usually based on a justifiable and testable conceptual framework, the demarcation of supraspecific boundaries is less objective and often subject to differences of opinion. The increased availability of large-scale phylogenies has in part promulgated a practice of what we consider excessively splitting clades at the “genus” level. Many of these new genus-level splits are predicated on untenable supporting evidence (e.g., weakly supported phylogenies and purportedly “diagnostic” but actually variable, non-exclusive, or otherwise problematic opposing character state differences) without careful consideration of the effects on downstream applications. As case studies, we critically evaluate several recent examples of splitting established monophyletic genera in four amphibian families that resulted in the creation/elevation of 20 genus-level names (Dicroglossidae: Phrynoglossus, Oreobatrachus, Frethia split from Occidozyga; Microhylidae: Nanohyla split from Microhyla; Ranidae: Abavorana, Amnirana, Chalcorana, Humerana, Hydrophylax, Indosylvirana, Papurana, Pulchrana, Sylvirana split from Hylarana; Rhacophoridae: Tamixalus, Vampyrius, Leptomantis, Zhangixalus split from Rhacophorus, Rohanixalus split from Feihyla, Orixalus split from Gracixalus, and Taruga split from Polypedates), and also address the taxonomic status of the monotypic genus Pterorana relative to Hylarana. We reassess the original claims of diagnosability and justifications for splitting and argue that in many cases, the generic splitting of clades is not only unnecessary but also destabilizes amphibian taxonomy, leading to a host of downstream issues that affect categories of the user community (stakeholders such as taxonomists, conservationists, evolutionary biologists, biogeographers, museum curators, educators, and the lay public). As an alternative, we advocate for the use of the subgenus rank in some cases, which can be implemented to establish informative partitions for future research without compromising on information content, while avoiding gratuitous (and often transient) large-scale binomial (genus-species couplet) rearrangements. We encourage taxonomists to consider the actual needs and interests of the larger non-taxonomic end-user community who fund the majority of taxonomic research, and who require a system that remains reasonably stable and is relatively intuitive, without the need for inaccessible laboratory equipment or advanced technical scientific knowledge to identify amphibian species to the genus level.