Carleton and Musser (1984) diagnosed the subfamily using morphological traits and reviewed general characters, major fossil groups, and past association with New World sigmodontines. Living hamsters form a monophyletic group bounded by unambiguously derived morphological traits. Their cladistic integrity is reinforced by phylogenetic analysis of nuclear gene sequences (Michaux and Catzeflis, 2000; Michaux et al., 2001b), which associate hamsters as a monophyletic clade within a major lineage that includes Neotominae and Arvicolinae (and Myospalacinae, but see that subfamily), apart from other muroid groups.

Corbet (1978c:88) noted that "generic divisions within the group are rather unstable and a fresh, comprehensive classification is required," a conclusion we echo after studying specimens and literature. One of these problematic divisions is Cricetulus, and Corbet (1978c:90) acknowledged that some species are frequently placed in Allocricetulus and Tscherskia, but "pending a review of generic classification in the subfamily as a whole," he preferred to include those taxa in Cricetulus. The highly distinctive and diagnosable morphologies of these forms as much argue the opposite: to retain them as separate genera until a systematic revision is performed, credible relationships are established, and genus-group synonymies are demonstrated. In cranial traits, Allocricetulus closely resembles Cricetus, not Cricetulus, and the latter is more like Phodopus than either Tscherskia or Allocricetulus (specimens studied in AMNH, FMNH, MCZ, and USNM). Gromov and Erbajeva (1995), Pavlinov and Rossolimo (1987, 1998), Pavlinov et al. (1995a), and Wang (2003) also treated Allocricetulus, and Tscherskia as genera separate from Cricetulus. Broad regional treatises condense much systematic and biological information on hamsters, including identification keys, in the Palearctic (Corbet, 1978c, 1984; Ellerman and Morrison-Scott, 1951), Europe (Mitchell-Jones et al., 1999; Niethammer, 1982a, d), Russia (Gromov and Erbajeva, 1995), China (Wang, 2003), and Indomalayan region (Corbet and Hill, 1992). Chromosomal information for European species recorded by Zima and Kral (1984a). Zagorodnyuk (1992b) reviewed taxonomic status of scientific names associated with samples from the Ukraine.

Antiquity of Cricetinae extends to the middle Miocene of N Africa, early Miocene of Europe, late Miocene of Asia, and Pliocene in the Mediterranean region. The seven extant genera are a remnant of a Neogene radiation resulting in a plethora of species in 16 extinct genera (extracted from Fahlbusch, 1996; Kälin, 1999; Kowalski, 2001; McKenna and Bell, 1997).