Lundomys molitor, also known as Lund's amphibious rat or the greater marsh rat, is a semiaquatic rat species from southeastern South America.

Its distribution is now restricted to Uruguay and nearby Rio Grande do Sul, Brazil, but it previously ranged northward into Minas Gerais, Brazil, and southward into eastern Argentina. The Argentine form may have been distinct from the living form from Brazil and Uruguay. L. molitor is a large rodent, with the head and body length averaging 193 mm (7.6 in), characterized by a long tail, large hindfeet, and long and dense fur. It builds nests above the water, supported by reeds, and it is not currently threatened.

Its external morphology is similar to that of Holochilus brasiliensis, and over the course of its complex taxonomic history it has been confused with that species, but other features support its placement in a distinct genus, Lundomys. Within the family Cricetidae and subfamily Sigmodontinae, it is a member of a group of specialized oryzomyine rodents that also includes Holochilus, Noronhomys, Carletonomys, and Pseudoryzomys.

Lundomys molitor was first described in 1888 by Danish zoologist Herluf Winge, who reviewed the materials Peter Wilhelm Lund had collected in the caves of Lagoa Santa, Minas Gerais, Brazil. Winge used four specimens for his description, including two skull fragments and an isolated maxilla (upper jaw) from the cave chamber Lapa da Escrivania Nr. 5 and a mandible (lower jaw) from Lapa da Serra das Abelhas, but the latter later turned out to be from a different species, probably Gyldenstolpia fronto. Lund named the animal Hesperomys molitor and placed it in the same genus (Hesperomys) as what is now Pseudoryzomys simplex and two species of Calomys. Subsequently, it was rarely mentioned in the literature on South American rodents; those authors who did mention it placed it in either Oryzomys or Calomys.

In 1926, American zoologist Colin Campbell Sanborn collected some rodents in Uruguay, which he identified as Holochilus vulpinus (currently Holochilus brasiliensis) in his 1929 report on the collection. When his successor at the Field Museum of Natural History, Philip Hershkovitz, reviewed Holochilus in 1955, he recognized that the series from Uruguay contained two species, one close to the forms of Holochilus found across much of South America, and another unique to Uruguay and southern Brazil; he named the latter as a new species, Holochilus magnus. Hershkovitz identified Holochilus as one of the members of a "sigmodont" group of American rodents, also including Sigmodon, Reithrodon, and Neotomys, on the basis of its flat-crowned molars, which are lophodont (the crown consists of transverse ridges). In 1981, H. magnus was also recognized in the Late Pleistocene of Buenos Aires Province, Argentina, and in 1982 it was recorded from Rio Grande do Sul in southern Brazil.

In a 1980 article, Argentine zoologist Elio Massoia recognized the resemblance between Winge's Hesperomys molitor and Hershkovitz's Holochilus magnus, and recommended that the former be reclassified as a species of Holochilus, Holochilus molitor. When American zoologists Voss and Carleton restudied Winge's material in a 1993 paper, they were unable to find any consistent differences between the two and accordingly considered them to pertain to the same species. In addition, they reviewed the differences between this species and other Holochilus and concluded that these were significant enough to place the former in a distinct genus, which they named Lundomys after Lund, who had collected the original material. Since then, the species has been known as Lundomys molitor.

In the same paper in which they described Lundomys, Voss and Carleton also, for the first time, diagnosed the tribe Oryzomyini in a phylogenetically valid way. Previously, Oryzomyini had been a somewhat loosely defined group defined among others by a long palate and the presence of a crest known as the mesoloph on the upper molars and mesolophid on the lower molars; this crest is absent or reduced in Holochilus and Lundomys. Voss and Carleton recognized five synapomorphies for the group, all of which are shared by Lundomys; the placement in Oryzomyini of Lundomys and of three other genera—Holochilus, Pseudoryzomys, and Zygodontomys—which also lack complete mesoloph(id)s has been universally supported since.

Voss and Carleton had found some support for a close relationship between Holochilus, Lundomys, and Pseudoryzomys within Oryzomyini. In subsequent years, the related species Holochilus primigenus and Noronhomys vespuccii were discovered, providing additional evidence for this grouping. The allocation of the former, which is similar to Lundomys in features of the dentition, to Holochilus is controversial, and placement as a second species of Lundomys has been suggested as an alternative. A comprehensive phylogenetic analysis of oryzomyines by Marcelo Weksler, published in 2006, supported a close relationship among Lundomys, Holochilus, and Pseudoryzomys; the other species of the group were not included. Data from the sequence of the IRBP gene supported a closer relationship between Holochilus and Pseudoryzomys, with Lundomys more distantly related, but morphological data placed Holochilus and Lundomys closer together, as did the combined analysis of both morphological and IRPB data. Subsequently, Carletonomys cailoi was described as an additional relative of Holochilus and Lundomys.