Abstract: The multiple correspondences between word classes and syntactic functions in Chinese pose a great challenge to the definition and classification of word classes, leading to debates over whether Chinese word classes pertain to the noun-verb inclusion or noun-verb division model. From the cognitive grammar perspective, word classes, e.g., nouns, verbs and adjectives reside in profiling, rooted in different cognitive operations. The match or mismatch between word classes and syntactic functions is derived from the categorization (i.e., instantiation and extension) of words by the functional position in relevant grammatical constructions, particularly low-level constructions. The correspondences between the verb phrase and subject/object, the noun phrase and predicate, as well as the conversions triggered by “XP de YP” or “hen X” are all motivated by these two types of categorization. Unlike English, Chinese prefers implicit coding strategies over explicit ones, leading to a significant number of words that belong to multiple word classes. Contrary to the word class-centered view presupposed by noun-verb inclusion and noun-verb division models, the construction-centered view posits that the class of a word is derived from the categorization of the grammatical construction it enters, and the phenomenon of one word belonging to more than one class poses no theoretical problem for Chinese word classes.