The epithelial sodium channel (ENaC) plays a key role in salt and water homeostasis in tetrapod vertebrates. There are four ENaC subunits (α, β, γ, δ), forming heterotrimeric αβγ- or δβγ-ENaCs. Although the physiology of αβγ-ENaC is well understood, for decades the field has stalled with respect to δβγ-ENaC due to the lack of mammalian model organisms. The SCNN1D gene coding for δ-ENaC was previously believed to be absent in rodents, hindering studies using standard laboratory animals. We analyzed all currently available rodent genomes and discovered that SCNN1D is present in rodents but was independently lost in five rodent lineages, including the Muridae (mice and rats). The independent loss of SCNN1D in rodent lineages may be constrained by phylogeny and taxon-specific adaptation to dry habitats, however habitat aridity does not provide a selection pressure for maintenance of SCNN1D across Rodentia. A fusion of two exons coding for a structurally flexible region in the extracellular domain of δ-ENaC appeared in the Hystricognathi (a group that includes guinea pigs). This conserved pattern evolved at least 41 Ma and represents a new autapomorphic feature for this clade. Exon fusion does not impair functionality of guinea pig (Cavia porcellus) δβγ-ENaC expressed in Xenopus oocytes. Electrophysiological characterization at the whole-cell and single-channel level revealed conserved biophysical features and mechanisms controlling guinea pig αβγ- and δβγ-ENaC function as compared with human orthologs. Guinea pigs therefore represent commercially available mammalian model animals that will help shed light on the physiological function of δ-ENaC.