The Chinese numeral, which was acknowledged as a separate part of speech approximately half a century ago, requires a theoretical insight. In this regard the task of prime importance is to distinguish between such key notions as “digit”, “number” and “numeral”. Confusing these notions may lead to fallacious deductions. The research is to be based on three languages: Russian, English and Chinese. At the beginning the notions under consideration are distinguished in Russian. To fulfill this task the author used lexicographical sources (defining dictionaries) as well as philosophical and linguistics dictionaries. As a result, in simple words, the triad “цифра - число - числительное” (“digit - number - numeral”) is interpreted as “symbol - notion - word”. Moreover, considering the peculiarities of the Chinese numeral as a part of speech, it was defined as a class of words performing a quantitative function and denoting an abstract number, the quantity of things, or mental categories of order of count, multiplicity, summation thatdeal with number. Then, by way of sememic analysis, the semantemes belonging to the Russian, English and Chinese lexemes were compared. The Russian-Chinese comparison revealed the following one-to-one correspondences: цифра - 数字, число - 数目, числительное - 数词. When comparing these results with the English corresponding variants, the lexeme “цифра” reveals a manifold synonymy in English: “figure”, “digit”, “cipher”, “numeral”, “number”. Thus the notion is united on the basis of the “symbol” sememe. Two counterparts are aligned with the lexeme “число”: “number” and “figure”. The lexeme “числительное” has unambiguous correlation in English and Chinese: “numeral” and “数词”, respectively. Semantic combinability is the basis for terminological synonymy which is observed in the English variant of the triad “цифра - число - числительное” (“digit - number - numeral”).