Phylogenetic allocation of Mystromys so puzzled Ellerman (1941:445) that he wrote "I am entirely at a loss to suggest the relationships of this genus, which seems not only isolated from the Palaearctic and Neotropical genera, but to have no marked generic characters." Based on a comprehensive study of morphological characters, Vorontsov (1966) concluded that M. albicaudatus is not closely related to Palearctic hamsters and placed it in the monotypic tribe Mystromyini. Carleton and Musser (1984:313) remarked that Mystromys is possibly a survivor of an ancient phyletic line, pointing out that Lavocat (1973, 1978) "raised such a novel possibility for Mystromys by noting its probable derivation from afrocricetodontine rodents and by recognizing the subfamily Mystromyinae in the Nesomyidae, a family composed of archaic African cricetids that he derived from the afrocricetodontines, the African counterpart to European and Asian cricetodontines." Pocock (1987), Skinner and Smithers (1990), and Denys (1991) retained Mystromys in the Cricetinae, family Cricetidae. Wessels (1996), based upon molar patterns, regarded Mystromys and Calomyscus as members of Myocricetodontinae, which she later arranged within Gerbillidae (Wessels, 1998), an alliance endorsed by Mein et al. (2000b).

No other morphological data suggest such a close relationship with either gerbils or Calomyscus (Vorontsov, 1966), and molecular data also contradict either allocation. In a phylogenetic analysis of two nuclear protein-coding genes (Michaux et al., 2001b), Mystromys is arrayed with representatives of Madagascan Nesomyinae and African Cricetomyinae and Dendromurinae, a clade completely separate from that containing Calomyscus and another circumscribing Gerbillinae, Deomyinae, and Murinae. According to the mitochondrial gene cytochrome b, Mystromys is similarly cladistically removed from Calomyscus and instead is sister genus to Petromyscus (Jansa et al., 1999). The ancestral stock of Mystromys and its close relative Proodontomys is yet to be identified, although Tong and Jaeger (1993) speculated that Mystromys was derived from Democricetodon, identified from the middle Miocene of East Africa. Mein et al. (2000b), however, disputed that origin and hypothesized that Mioharimys, late Miocene of Namibia, represents the ancestral group from which living Mystromys evolved.