Oryzomyini. Arranged by Hershkovitz (1955a) as one of four genera of sigmodont rodents. Based on reproductive anatomy, Hooper and Musser (1964) remarked that Holochilus may represent a "well differentiated oryzomyine rather than a sigmodont." Retained, with Sigmodon proper, in the tribe Sigmodontini by Reig (1984, 1986); transferred to Oryzomyini by Voss and Carleton (1993), who associated Holochilus within a tetralophodont clade that includes Pseudoryzomys and Lundomys (also see Carleton and Olson, 1999). Membership in Oryzomyini sustained by taxonomically broad, cladistic studies of morphological, karyological, and molecular characters (Baker et al., 1983; Malygin and Rosmiarek, 1997; Smith and Patton, 1999; Steppan, 1995; Weksler, 2003).

We follow the proposal of Voss and Abramson (1999) for designating H. sciureus Wagner, 1842, as the type species of Holochilus in order to continue traditional usage of the genus for a myomorphous sigmodontine. As they explain, the type species as heretofore given, Mus (Holochilus) leucogaster Brandt, 1835, was regrettably listed by Miller and Rehn (1901:89) without specimen examination and is based on an example of the echimyid Trinomys. Mus simpsonii Philippi, 1900, occasionally listed as a species of Holochilus (e.g., Gyldenstolpe, 1932), was reidentified as an example of Rattus norvegicus by Osgood (1943). See Musser et al. (1998:280-284) for the contorted taxonomic history of Mus physodes Brants, 1827 (or Lichtenstein, 1827), formerly associated with Holochilus, whether as species (Ellerman, 1941; Gyldenstolpe, 1932) or as junior synonym of H. brasiliensis (Cabrera, 1961; Hershkovitz, 1955a).

Genus revised by Hershkovitz (1955a), who consolidated 13 nominal species (e.g., Ellerman, 1941) under H. brasiliensis and diagnosed a new one, H. magnus; the latter has been removed to the genus Lundomys (Voss and Carleton, 1993). Subsequent studies have revealed that brasiliensis of Hershkovitz (1955a) is a composite of three or more species (Aguilera and Pérez-Zapata, 1989; Gardner and Patton, 1976; Massoia, 1980a, 1981; Reig, 1986). In general, evidence for the species acknowledged below issues from more localized studies that lack the persuasion of a full generic review; these names, distributions, and synonyms must therefore be accepted as tentative pending such a synoptic revision. Our association of synonyms basically observes the species classification of Massoia (1980a). Known from the middle Pleistocene, Bolivia, based on the fossil species, H. primigeneus, which Steppan (1996) described as possibly ancestral to living Holochilus; also early through late Pleistocene of Argentina, as H. brasiliensis (Pardiñas, 1999).