Toponymic evidence suggests Ainu was once spoken in northern Honshu and that much of the historically attested extent of the family was due to a relatively recent expansion northward.

Hattori (1964) considered Ainu data from 19 regions of Hokkaidō and Sakhalin, and found the primary division to lie between the two islands.

Hokkaidō Ainu clustered into several dialects with substantial differences between them: the 'neck' of the island (Oshima County, data from Oshamambe and Yakumo); the "classical" Ainu of central Hokkaidō around Sapporo and the southern coast (Iburi and Hidaka counties, data from Horobetsu, Biratori, Nukkibetsu and Niikappu; historical records from Ishikari County and Sapporo show that these were similar); Samani (on the southeastern cape in Hidaka, but perhaps closest to the northeastern dialect); the northeast (data from Obihiro, Kushiro and Bihoro); the north-central dialect (Kamikawa County, data from Asahikawa and Nayoro) and Sōya (on the northwestern cape), which was closest of all Hokkaidō varieties to Sakhalin Ainu.

Scant data from Western voyages at the turn of the 19th–20th century (Tamura 2000) suggest there was also great diversity in northern Sakhalin, which was not sampled by Hattori.

A study by Lee and Hasegawa of Waseda University found evidence that the Ainu language and the early Ainu-speakers originated from the Northeast Asian/Okhotsk population, which established themselves in northern Hokkaido and expanded into large parts of Honshu and the Kurils.

Vovin (1992) presented his reconstruction of Proto-Ainu with evidence, in the form of proposed sound changes and cognates, of a relationship with Austroasiatic.

Linguists believe the vocabulary shared between Ainu and Nivkh (historically spoken in the northern half of Sakhalin and on the Asian mainland facing it) is due to borrowing.

An example is modern Japanese sake or shake (鮭), meaning 'salmon', probably from the Ainu sak ipe or shak embe for 'salmon', literally 'summer food'.

Asahi (2005) reported that the status of the Ainu language was rather high and was also used by early Russian and Japanese administrative officials to communicate with each other and with the indigenous people.

[20] It is occasionally suggested that Ainu was the language of the indigenous Emishi people of the northern part of the main Japanese island of Honshu.

[21] Other place names in Kantō and Chūbu, such as Mount Ashigara (Kanagawa–Shizuoka), Musashi (modern Tokyo), Keta Shrine (Toyama), and the Noto Peninsula, have no explanation in Japanese, but do in Ainu.

However, it has also been proposed that the Ainu themselves can be identified with the Okhotsk culture, and that they expanded south into northern Honshu as well as to the Kamchatka Peninsula,[8][25] or that the Emishi spoke a Japonic language, most closely related to ancient Izumo dialect, rather than anything related to Ainu, with Ainu-speakers migrating later from Hokkaido to northern Tōhoku.

The purported evidence for this are old-Japanese loanwords in the Ainu language, including basic vocabulary, as well as distinctive Japonic terms and toponyms found in Tōhoku and Hokkaido, that have been linked to the Izumo dialect.