How common are chengyu in everyday Chinese?
Common enough that you meet one every page or two, few enough that a few hundred do most of the work.
We make Mandarin The Wild.
How often they come up
We took the Xinhua idiom dictionary's list of chengyu and counted them in two kinds of Chinese: web text, and the speech in 292 films and shows made in Mandarin. chinese-xinhua: the Xinhua idiom dictionary’s list (30,895 entries), the chengyu we counted. FineWeb-2, the web text we counted (its Mandarin part, cmn_Hani). Lison and Tiedemann (2016), OpenSubtitles2016: extracting large parallel corpora from movie and TV subtitles, LREC 2016 (the subtitles we counted).
| Kind of Chinese | Chengyu per 1,000 words | Chengyu that make up half of all uses |
|---|---|---|
| Web text | 2.9 | 563 |
| Film and TV speech | 3.75 | 414 |
Speech uses slightly more, and a different set. The web's most common are 不好意思, 一如既往, 一模一样, 众所周知 and 莫名其妙. On screen they are 不好意思, 阿弥陀佛, 乱七八糟, 胡说八道 and 无论如何: complaints, oaths and exclamations rather than the set phrases of an article.
你别胡说八道！
to talk nonsense; complete rubbish
Also written 胡說八道
Most of the list never comes up
In 1.7 million words of film and TV speech, only 2,639 different chengyu appeared; more than 27,000 of the 30,219 on the list never came up once. A chengyu dictionary is mostly a record of literature and history, not of what people say.
Which ones to learn
The ones you will meet. But the web's list and the screen's list differ more than you might think:
| The web's most common chengyu you know | Share of chengyu uses on the web | Share in film and TV speech |
|---|---|---|
| 100 | 19% | 13% |
| 500 | 47% | 25% |
| 1,000 | 67% | 36% |
Knowing the web's top 1,000 covers two thirds of the chengyu you read on the web but only about a third of those you hear in films and shows. So learn chengyu from the Chinese you actually read and watch, one at a time, as you meet them.
Native speakers know the common ones first
In ratings of 350 chengyu by native speakers, how familiar a chengyu felt, how often people thought they met it and how early they had learned it all went together. Li, Zhang and Wang (2016), Descriptive norms for 350 Chinese idioms with seven syntactic structures, Behavior Research Methods 48(4). The common ones are the ones everyone knows, which is one more reason to start with what you meet.
What the HSK expects
The 2025 HSK list has 402 chengyu: 不好意思 at level 2, 十字路口 at level 4, four at level 5 (among them 丰富多彩 and 五颜六色), eleven at level 6 (among them 一模一样, 众所周知 and 人山人海) and 385 at levels 7–9. HSK: the official site of the Chinese Proficiency Test. So the exam leaves chengyu almost entirely to its advanced levels, even though web text and films use them from the start.
A plan that fits the evidence
- Learn the chengyu you meet, not a list. Point at one on any page or in a video's subtitles with Mandarin The Wild to see what it means as a whole and what each character adds, and save it to your Backpack.
- Start with the common ones. Chinese idioms lists them most common first, each with its pinyin, meaning and real sentences.
- Test yourself. The chengyu quiz plays ten chengyu from the HSK 7–9 list and asks what each one means.
How we counted
Chengyu are the entries of four characters or more on the Xinhua idiom dictionary's list. chinese-xinhua: the Xinhua idiom dictionary’s list (30,895 entries), the chengyu we counted. Web counts are our engine's counts of the Mandarin part of FineWeb-2, which stop at about 0.2 uses in a million words, so the rarest chengyu are missing there. FineWeb-2, the web text we counted (its Mandarin part, cmn_Hani). Film and TV speech is the subtitles of the 292 OpenSubtitles titles whose original language is Mandarin, cut into words by our engine. Lison and Tiedemann (2016), OpenSubtitles2016: extracting large parallel corpora from movie and TV subtitles, LREC 2016 (the subtitles we counted). HSK levels are the 2025 syllabus.
Questions
How common are chengyu in Chinese?
In our count, about 3 in every 1,000 words of Chinese web text and 3.75 in every 1,000 words of film and TV speech: roughly one every page or two.
How many chengyu should I learn?
A few hundred do most of the work: 563 make up half of all chengyu uses on the web and 414 in films and shows. Learn the ones you meet first.
Are chengyu on the HSK?
Mostly at the top. The 2025 HSK list has 402 chengyu, and 385 of them are at levels 7–9.
Sources
Checked on 14 October 2026. Prices and features change; tell us if one has.
- chinese-xinhua: the Xinhua idiom dictionary’s list (30,895 entries), the chengyu we counted
- Li, Zhang and Wang (2016), Descriptive norms for 350 Chinese idioms with seven syntactic structures, Behavior Research Methods 48(4)
- Lison and Tiedemann (2016), OpenSubtitles2016: extracting large parallel corpora from movie and TV subtitles, LREC 2016 (the subtitles we counted)
- FineWeb-2, the web text we counted (its Mandarin part, cmn_Hani)
- HSK: the official site of the Chinese Proficiency Test