Kari and Potter (2010:10) place the total territory of the 53 Athabaskan languages at 4,022,000 square kilometres (1,553,000 sq mi).

[5] Linguists conventionally divide the Athabaskan family into three groups, based on geographic distribution: The 38 Northern Athabaskan languages are spoken throughout the interior of Alaska and the interior of northwestern Canada in the Yukon and Northwest Territories, as well as in the provinces of British Columbia, Alberta, Saskatchewan and Manitoba.

Eyak and Athabaskan together form a genealogical linguistic grouping called Athabaskan–Eyak (AE) – well-demonstrated through consistent sound correspondences, extensive shared vocabulary, and cross-linguistically unique homologies in both verb and noun morphology.

With Jeff Leer's 2010 advances, the reconstructions of Na-Dene (or Athabascan–Eyak–Tlingit) consonants, this latter grouping is considered by Alaskan linguists to be a well-demonstrated family.

The Alaska Native Language Center, for example, takes the position that recent improved data on Haida have served to conclusively disprove the Haida-inclusion hypothesis.

Edward Vajda of Western Washington University summarized ten years of research, based on verbal morphology and reconstructions of the proto-languages, indicating that these languages might be related.

[8] The internal structure of the Athabaskan language family is complex, and its exact shape is still a hotly debated issue among experts.

The conventional three-way split into Northern, Pacific Coast, and Southern is essentially based on geography and the physical distribution of Athabaskan peoples rather than sound linguistic comparisons.

Kwalhioqua–Clatskanai (#7) was normally placed inside the Pacific Coast grouping, but a recent consideration by Krauss (2005) does not find it very similar to these languages.

As Tuttle and Hargus put it, "we do not consider the points of difference between the two models ... to be decisively settled and in fact expect them to be debated for some time to come."

Due to the failure of the usual criteria of shared innovation and systematic phonetic correspondences to provide well-defined subgroupings, the Athabaskan family – especially the Northern group – has been called a "cohesive complex" by Michael Krauss (1973, 1982).

Athabaskanists have concluded that it is a Northern Athabaskan language consistent with its geographical occurrence, and that it might have some relation to its distant neighbor Tahltan.

There are many homologies between Proto-Athabaskan vocabulary and patterns reflected in archaeological sites such as Upward Sun, Swan Point and Broken Mammoth (Kari 2010).

Using computational phylogenetic methods, Sicoli & Holton (2014)[9] proposed the following classification for the Athabaskan languages based exclusively on typological (non-lexical) data.