In the Southeast Asian Sinosphere, we can observe a circulation of ethnonyms between local languages and Chinese. A local Southeast Asian autonym is borrowed into Chinese, undergoes sound changes affecting the Chinese language, and then returns to the original populations, or to other populations. The result is a coexistence of ethnonyms of highly different phonetic outlook but originating in the same etymon. We will examine four families of ethnonyms, mostly Austroasiatic and Tai-Kadai. The first observation that emerges from this survey is that there is considerable phonetic variety within each family of ethnonyms. Among the four fundamental etymons, only *k.riː has been fully preserved until today, coexisting with its daughter forms Hlai/ Li and Tai/ Thai which name Tai-Kadai populations (leaving aside the dubious case of ‘Yi’). The use of the daughter forms of *k.raw is common to Tai-Kadai (Gelao, Lao) and to Austroasiatic ancestors of the Vietnamese. The etymon *p.rak/ b.rak is mainly associated with Austroasiatic populations, and secondarily with Tibeto-Burman ethnic groups. It can be hypothsized that this situation reflects a spread of Tibeto-Burman languages over an Austroasiatic substratum. The etymon *-raŋ is shared between the Asian mainland and the Western Indonesian world. It is attested today in three linguistic families: Tibeto-Burman, Austroasiatic and Austronesian.