This paper proposes a tripartite model describing the lexical categories across different registers and levels of formality in the Cantonese language in contemporary Hong Kong: (1) native Cantonese words; (2) Sino-Cantonese words, and (3) Anglo-Cantonese words. Examples of authentic Cantonese use were used to illustrate the histories and etymology of key lexical categories and sub-categories as found in the city's linguistic landscape. As a sensitising device, the proposed classificatory model highlights the role of lexical borrowings in the constitution of contemporary Cantonese lexis, whilst decentring a primarily Mandarin-based approach to research and practice. Given the authenticity and omnipresence of Cantonese use across spoken and written modalities in contemporary Hong Kong, this paper argues that there is much scope for disambiguating and systematising the place of Cantonese lexis in the local Chinese language curriculum. In this regard, the case of Chinese language provision for ethnolinguistic minority learners with Chinese-as-an-Additional-Language (CAL) needs in post-handover Hong Kong is put forth to call attention to the utility of this descriptive model in mitigating against the learning and pedagogical issues associated with the disconnect between the curriculum and authentic language use, as well as linguistic disintegration.