In the openly licensed Chinese text we parsed, 个 (個, gè) filled 47% of the places where a measure word counts a noun. In essays by learners of Chinese it filled 68%. The rest went to specific measure words (量词), also called classifiers, such as 本 for books and 张 (張) for flat things, and fewer than two dozen of those did most of the work.
The text came from three collections in which every word's grammatical role has been tagged by hand, published by the Universal Dependencies project: Chinese Wikipedia, news and Wikipedia sentences translated into Chinese from European languages, and written Chinese versions of Hong Kong film subtitles and a Legislative Council record. Together they hold 691 such places. The learner sample is 122 places in 451 sentences from learners' essays, at a level the collection does not state. We left out time words, units such as 公斤 and 元, and measure words for actions such as 次. Chinese here is Simplified, with the Traditional form in brackets where it differs.
Few classifiers in our sample fall outside a core of 24
The three samples used 71 different measure words for nouns. The ten most frequent covered 77.6% of all uses, and 24 of them, 个 included, covered 90%. After 个, the most frequent were 种 (種, kinds), 位 (people, politely), 名 (people, in counts), 项 (項, items), 座 (buildings and mountains), 场 (場, events), 家 (businesses), 部 (films and works), 张 (張), 条 (條, long things), 所 (institutions), 支 (teams and long thin things), 件 (matters and garments) and 份 (portions and copies).
That 24 is a count from our sample, not a rule of the language. One ranking guide says 个 "represents 94% of measure word usage in everyday speech", and gives no source. In each of our three samples it was between 38% and 51%: 47.1% of 244 places in Wikipedia, 50.6% of 332 in the translated sentences and 38.3% of 115 in the Hong Kong text. The likely range for the pooled figure is 44% to 51%.
Nearly all the classifiers we counted appear in the syllabus by level 4
The HSK exam syllabus published in November 2025 and in force since July 2026 marks 160 distinct words with 量, measure word, as a main part of speech, across levels 1 to 6 and a combined band for 7 to 9. Level 1 has 18, and by level 3 there are 43. A figure now repeated in search summaries puts 28 measure words in the first three levels; we found no source for it. Our guide to HSK word counts by level covers the rest of that list. The exam is mainland China's. Taiwan tests learners with TOCFL, and school Chinese in Singapore and Malaysia follows its own curricula, which our HSK vocabulary lookup also points out.
Measured against the list, 91% of the measure words used in our three samples are words the syllabus has introduced by level 4 of the new nine-level scheme, counting a word whatever part of speech the syllabus gives it. By level 1 it is 53%, mostly because 个 is a level 1 word. Two gaps stand out. 把, for chairs, umbrellas and knives, is a level 3 grammar point, and 种 is the most frequent classifier after 个 in our samples.
Until 30 September 2026, our measure word picker had no noun counted with either, and its Taiwan page printed 一只猫 with the mainland 只 where Taiwan writes 隻. It now lists 97 nouns, including rows for 把 and 种, in each place's own script.
Most of the learners' classifiers are 个
In the learners' essays, 个 filled 68% of the places, with a likely range of 59% to 76%. They used 25 different measure words, 13 of them covered 90% of uses, and 79% of their measure words were level 1 words. Learners also put 个 on 餐厅 (餐廳, restaurants) and 电话 (電話, phone calls), where the Taiwan dictionary described below counts shops with 家 and phone calls with 通.
Part of the gap may be genre, since an encyclopedia and a personal essay count different things. When we also left out words for kinds and groups, such as 种 and 群, the gap held: 个 was 52% of the rest in the three samples and 70% in the essays.
About half the nouns in our dictionary check take several classifiers
Taiwan's Ministry of Education dictionary, 《重編國語辭典修訂本》, gives example phrases under each measure word. We checked 71 everyday nouns against it. Of the 35 it cites under any measure word, 17 appear under two or more. Fish take 条 (條) and also 尾, as in 「一尾魚」. People take 个, 位, 名 and 口, and the dictionary's note on 位 is 「多含敬意」, usually respectful, which is a matter of tone rather than a requirement.
Most entries define a measure word by a class of thing rather than a list. 條 is 「計算條狀物的單位」, the unit for long, strip-shaped things. 頭 counts cattle, donkeys, mules, sheep and pigs, and also heads of garlic. 匹 counts horses, mules and donkeys, and bolts of cloth. A noun that fits the class can take the word even when no dictionary example names it.
Some classifiers are cups and bowls
The syllabus's grammar list splits measure words for nouns into dedicated ones, such as 本 and 张, and borrowed ones: 杯 at level 1, 包 at level 2, and 碗 and 盘 (盤) at level 3. A borrowed measure word is a container, so water is counted as 一杯水 or 一碗水, a cup or a bowl of it. The same list also puts 天, 年 and 岁 (歲) among its level 1 measure words, which is why 三天, three days, and 三年, three years, need nothing between the number and the word.
Several nouns take 个 and a specific classifier in the same text
In the three samples, 公司 (company) took 家 four times, 个 twice and 间 (間) once. 人 took 个 fourteen times and 位 twice, and 系统 (系統, system) took 个 four times and 套 once. The same writers used both, so 个 and a specific word are often alternatives rather than right and wrong.
With some nouns 个 is not a choice. 一个月 is a month, and 一月 is January, and 19 of the 327 uses of 个 in the three samples were counting a month.
View author profile → · Editorial policy
