The paper deals with measure words (numeral classifiers) in the Indonesian language — a special category of service tokens used for measurement (possibility of counting) of mentioned things and pointing to the noun belonging to a particular semantic class (category). Such words exist in many languages of East and South-East Asia (Vietnamese, Indonesian, Chinese, Korean, Malay, Japanese, etc.), as well as Indo-Iranian, Turk, Dravidian languages. We identified and described these categories, marked by a number of characteristic features of the use of such tokens in the Indonesian language, and made some parallels with the Russian language, which also has such category of words. The author examines about 30 measure words of the Indonesian language, describes their their category, notes a number of characteristic features of using such lexemes in the Indonesian language, makes some parallels with the Russian language, which also has the similar words. He also notes that if a category of countability is undeveloped it is measure words that indicate possibility of counting (measurement) of sujects, which explains the need for them in the Indonesian language. It is also indicated that the classifiers play an important substantiating role, i.e. they point on moving of other parts of speech into the category of nouns (moreover they play also emphasizing role that resembles the function of articles in the English language). Having substantiating role such measure words also provide referential attitude of a numeral to a specific noun, regardless of whether the noun itself is named or not.