The 100 most common Chinese characters

The most common Chinese characters in spoken Mandarin are 我 wǒ (I), 的 de (the particle that links a describing word to a noun), 你 nǐ (you), 是 shì (to be), 了 le (the particle for a finished action), 不 bù (not), 一 yī (one), 们 men (the plural ending for people), 这 zhè (this) and 他 tā (he). We counted every character in 122 million characters of Chinese film and television dialogue: those ten make up 21.7% of it, and the hundred in the table below make up 54.3%. The table gives each one's pinyin, its meaning, how often it stands alone as a word, and the words it is most often part of.

Updated 10 October 2026 · 7 min read

How the characters were counted

The source is the Chinese (simplified) list from FrequencyWords, built from the OpenSubtitles 2018 corpus of film and television subtitles. That list counts words, because the corpus was split into words before counting. To count characters, every word was broken into its characters and each character was credited with the word's count, once for each time it appears: 我们 wǒmen (we), said 1,000 times, adds 1,000 to 我 and 1,000 to 们. In total the dialogue holds 11,302 different characters.

Two columns in the table are worth explaining before you read it:

Alone is the share of a character's appearances in which it is a whole word on its own. 吗 ma is 98%: it is almost never part of a longer word. 么 me is 3%: it is almost always the second half of 什么 or 怎么.
Rank in writing is the character's place in Jun Da's frequency list of modern written Chinese, about 194 million characters of text collected online. It shows how far speech and writing part company.

The pinyin is the reading the character has in most of its uses. Where it has others, the note under the meaning gives them. Every reading and meaning was checked against the CC-CEDICT dictionary and against our own Mandarin word list.

The 100 most common Chinese characters

#	Character	Pinyin	Meaning	Per million characters	Alone	Commonest words with it	Rank in writing
1	我	wǒ	I, me	41,106	73%	我们、我要、我会	9
2	的	de	particle linking a describer to a noun; 's, of dí in 的确 (indeed), dì in 目的 (aim)	34,762	94%	—	1
3	你	nǐ	you	29,770	91%	—	32
4	是	shì	to be; yes, that's right	23,664	58%	不是、就是、只是	3
5	了	le	particle: action completed, or a change of state liǎo in 了解 (to understand) and in 受不了 (can't stand)	19,327	91%	—	5
6	不	bù	not, no said bú before a fourth tone: 不是 búshì	18,986	31%	不是、不会、不能	4
7	一	yī	one; a said yí before a fourth tone (一个 yí ge) and yì before the others (一起 yìqǐ)	13,116	4%	一个、一起、一下	2
8	们	men	plural ending for people never a word on its own in normal use: 我们, 他们, 你们	12,478	4%	我们、他们、你们	13
9	这	zhè	this often zhèi in speech	12,265	30%	这个、这里、这样	11
10	他	tā	he, him	11,387	65%	他们、他妈的、其他	10
11	在	zài	at, in; to be (somewhere); in the middle of (doing)	11,330	74%	现在、正在、存在	6
12	有	yǒu	to have; there is	11,129	38%	没有、还有、所有	8
13	么	me	ending of 什么 (what), 怎么 (how), 这么 (so) almost never on its own	9,881	3%	什么、怎么、为什么	63
14	好	hǎo	good, well; OK hào, to be fond of, in 爱好 (hobby)	8,405	55%	你好、最好、好像	82
15	人	rén	person, people	8,202	32%	有人、女人、男人	7
16	个	gè	the general measure word for a noun with no measure word of its own; toneless in 一个 (yí ge)	7,788	16%	一个、这个、那个	12
17	要	yào	to want; will; must yāo, to demand, in 要求	7,712	38%	不要、需要、我要	26
18	来	lái	to come	7,347	31%	回来、起来、出来	15
19	那	nà	that often nèi in speech	6,739	39%	那个、那么、那些	38
20	什	shén	first half of 什么 shénme, what almost only in 什么	6,594	1%	什么、为什么、干什么	156
21	就	jiù	then; just, exactly; right away	6,202	67%	就是、就要、就算	27
22	会	huì	can (know how to); will; meeting	6,068	42%	不会、我会、机会	29
23	到	dào	to arrive; to, until	6,048	27%	看到、找到、得到	22
24	没	méi	not (have), didn't méi negates 有 and the past; 不 negates the rest	5,526	25%	没有、没事、没错	72
25	吗	ma	particle that makes a yes-no question	5,500	98%	—	453
26	去	qù	to go	5,081	60%	出去、过去、回去	64
27	说	shuō	to say, to speak	5,002	68%	说话、来说、听说	24
28	想	xiǎng	to want to; to think; to miss	4,789	54%	不想、想要、想法	99
29	上	shàng	up, on, above; to go up; last (week)	4,590	29%	上帝、马上、晚上	16
30	为	wèi	for, because of (为什么, why) wéi, to be, to act as, in 认为 (to think) and 以为 (to assume)	4,481	14%	为什么、因为、认为	18
31	能	néng	can, to be able to	4,439	34%	不能、可能、我能	35
32	道	dào	road, way; to say almost only inside 知道 (to know)	4,398	1%	知道、难道、道歉	52
33	看	kàn	to look, to watch, to read	4,349	27%	看到、看看、看起来	76
34	吧	ba	particle for suggestions (let's …) and guesses (… right?)	4,320	96%	—	470
35	可	kě	can, may (可以); but (可是)	4,306	10%	可以、可能、可是	30
36	她	tā	she, her said exactly like 他, he	4,165	96%	—	91
37	里	lǐ	inside, in in 这里 (here), 那里 (there), 哪里 (where)	4,082	18%	这里、那里、哪里	50
38	很	hěn	very	4,032	75%	很多、很快、很棒	138
39	知	zhī	to know almost only inside 知道	4,025	0%	知道、通知、不知	123
40	得	de	particle after a verb: how, or how well (跑得快, runs fast) dé, to get, in 得到; děi, must; which of the three is commonest in speech is our call, not a measurement	4,000	32%	觉得、得到、记得	39
41	以	yǐ	by means of; in order to mostly in 可以 (can) and 所以 (so)	3,916	5%	可以、所以、以为	23
42	子	zi	noun ending (孩子 child, 房子 house) zǐ, son, child, when it carries meaning	3,784	1%	孩子、儿子、房子	37
43	事	shì	matter, thing, business	3,727	30%	事情、没事、怎么回事	58
44	都	dōu	all, both; already dū, capital city, in 首都	3,588	95%	—	68
45	生	shēng	to be born; life; raw in 先生 (Mr), 发生 (to happen), 医生 (doctor)	3,505	2%	先生、发生、生活	34
46	下	xià	down, under; next; to go down	3,387	13%	一下、下来、下去	42
47	过	guò	to pass, to cross; after a verb: have ever (去过, have been)	3,386	31%	不过、过来、过去	46
48	对	duì	right, correct; towards, to	3,384	63%	对不起、绝对、对于	33
49	也	yě	also, too	3,336	83%	也许、再也、也好	31
50	还	hái	still, yet; also huán, to give back	3,322	48%	还有、还是、还好	80
51	大	dà	big dài in 大夫 (doctor)	3,290	13%	大家、大概、大人	17
52	样	yàng	kind, way, look in 这样 (like this), 一样 (the same), 怎么样 (how)	3,030	0%	这样、一样、怎么样	88
53	啊	a	exclamation; softens a sentence	2,970	99%	—	753
54	时	shí	time; when in 时候 and 时间	2,861	12%	时候、时间、小时	25
55	真	zhēn	real, true; really	2,858	18%	真的、真是、真正	204
56	和	hé	and, with; peace	2,837	94%	—	19
57	天	tiān	day; sky in 今天 (today), 明天 (tomorrow)	2,784	14%	今天、明天、一天	78
58	做	zuò	to do, to make	2,733	77%	做到、做出、做好	246
59	怎	zěn	how only in 怎么 and 怎样	2,715	2%	怎么、怎么样、怎样	382
60	出	chū	to go out, to come out	2,635	9%	出来、出去、出现	28
61	起	qǐ	to rise, to get up in 一起 (together) and 对不起 (sorry)	2,563	4%	一起、起来、对不起	75
62	但	dàn	but	2,516	65%	但是、但愿、不但	95
63	开	kāi	to open; to start; to drive	2,482	6%	开始、离开、开心	94
64	家	jiā	home, family in 大家 (everyone)	2,466	7%	大家、回家、家伙	56
65	小	xiǎo	small, little, young	2,426	13%	小姐、小心、小时	83
66	只	zhǐ	only, just zhī, the measure word for animals and one of a pair	2,406	19%	只是、只有、只要	97
67	多	duō	many, much; more	2,392	29%	很多、多少、多久	61
68	点	diǎn	dot, point; a little (一点); o'clock	2,374	18%	一点、快点、有点	128
69	现	xiàn	now, present; to appear almost only in 现在 (now), 发现, 出现	2,338	0%	现在、发现、出现	70
70	给	gěi	to give; for, to	2,321	82%	交给、还给、嫁给	180
71	如	rú	like, as; if mostly in 如果 (if)	2,312	2%	如果、如何、如此	67
72	把	bǎ	moves the object before the verb (把门关上, shut the door)	2,294	93%	—	110
73	回	huí	to go back, to return; time (once, twice)	2,105	7%	回来、回家、回去	172
74	让	ràng	to let; to make (someone do)	2,087	93%	—	339
75	着	zhe	particle: an ongoing state (看着, looking at) zháo in 睡着 (asleep); zhuó in a few written words	2,082	53%	看着、活着、跟着	41
76	谢	xiè	to thank almost only in 谢谢	2,074	2%	谢谢、感谢、多谢	897
77	所	suǒ	place mostly in 所以 (so) and 所有 (all)	2,060	11%	所以、所有、所有人	54
78	意	yì	meaning, idea, wish in 意思 (meaning), 愿意 (willing), 注意 (to pay attention)	2,057	0%	意思、愿意、注意	104
79	些	xiē	some, a few in 这些 (these), 那些 (those), 一些 (some)	2,049	6%	这些、那些、一些	86
80	后	hòu	after, behind, back	2,040	17%	然后、最后、之后	48
81	走	zǒu	to walk; to leave	2,029	69%	走开、带走、拿走	207
82	自	zì	self; from mostly in 自己 (oneself)	2,007	2%	自己、自由、来自	43
83	然	rán	so, like that only in words: 然后 (then), 当然 (of course), 突然 (suddenly)	1,974	1%	然后、当然、虽然	55
84	别	bié	don't!; other	1,933	38%	别人、特别、别的	222
85	心	xīn	heart, mind	1,886	4%	担心、小心、开心	90
86	地	dì	earth, ground, place de, the particle that makes an adverb (慢慢地, slowly)	1,886	15%	地方、地球、地狱	21
87	斯	sī	spells the sound s in foreign names 詹姆斯 James, 克里斯 Chris	1,876	4%	詹姆斯、克里斯、罗斯	168
88	快	kuài	fast; soon	1,840	41%	快点、很快、快乐	366
89	它	tā	it said exactly like 他 and 她	1,749	80%	它们、其它、它会	107
90	果	guǒ	fruit; result almost only in 如果 (if) and 结果 (result)	1,741	0%	如果、结果、果然	165
91	话	huà	speech, words in 电话 (telephone), 说话 (to talk), 的话 (if)	1,729	8%	的话、电话、说话	170
92	用	yòng	to use; need (不用, no need)	1,719	34%	不用、使用、利用	51
93	死	sǐ	to die; dead	1,698	41%	该死、死亡、杀死	317
94	呢	ne	particle: and …?, what about …?	1,680	99%	—	383
95	跟	gēn	with; to follow	1,675	85%	跟着、跟我来、跟踪	541
96	再	zài	again; then in 再见 (goodbye)	1,671	53%	再见、不再、再说	242
97	打	dǎ	to hit; to make (a call) in 打电话 (to phone)	1,670	26%	打电话、打算、打开	223
98	定	dìng	fixed, settled; to decide in 一定 (certainly), 决定 (to decide)	1,665	1%	一定、肯定、决定	77
99	行	xíng	OK, all right (不行, no way); to walk háng, row, trade, in 银行 (bank)	1,663	12%	不行、进行、行动	53
100	太	tài	too (much); very	1,660	27%	太好了、太太、太棒了	240

"Per million characters" is how many times the character turns up in every million characters of dialogue. Where a character stands alone 90% of the time or more, the words column is left empty: the few longer words it appears in are not what it is for.

Characters that are almost never words

21 of the hundred stand alone less than 5% of the time. They are building blocks, and a learner meets them inside words rather than on their own:

么 me, 13th, and 什 shén, 20th, exist to make 什么 shénme, "what". 怎 zěn is there for 怎么 zěnme, "how".
知 zhī and 道 dào are 39th and 32nd almost entirely because of one word, 知道 zhīdào, "to know". 道 on its own is "road" or "way", a meaning that barely registers in dialogue.
现 xiàn is 现在 xiànzài, "now". 果 guǒ is "fruit" in the dictionary and 如果 rúguǒ, "if", in conversation. 然 rán is 然后 ránhòu (then) and 当然 dāngrán (of course). 意 yì is 意思 yìsi (meaning).

At the other end, 12 of the hundred are a whole word nearly every time: the sentence particles 吗, 吧, 呢 and 啊, the pronouns 你 and 她, and small, hard-working words such as 都 dōu (all), 让 ràng (to let) and 把 bǎ, which moves the object in front of the verb.

The practical lesson is that learning a character is not the same as learning a word. Knowing 知 and 道 separately does not tell you that together they mean "to know". The difference between counting characters and counting words, and why both counts are correct, is set out in how many words you need to speak Mandarin.

Spoken Chinese is not written Chinese

71 of our hundred are also in the top hundred of Jun Da's written list. The other 29 are not, and the differences show what conversation is made of.

Character	Rank in dialogue	Rank in writing
我 wǒ, I	1	9
的 de	2	1
你 nǐ, you	3	32
吗 ma, question particle	25	453
吧 ba, suggestion particle	34	470
啊 a, exclamation	53	753
中 zhōng, middle	113	14
国 guó, country	158	20

Speech is about you and me. 我 is first in dialogue and 你 third. In writing, 的 leads and 你 falls to the thirties.
Speech has particles that writing barely uses. 吗, 吧, 呢 and 啊 change a sentence's tone: a question, a suggestion, an "and you?", an exclamation. All four are in the spoken top hundred, and none of them is in the written top 300.
Writing has its own vocabulary. 国 is 20th in writing, the 国 of 中国 (China) and 国家 (country, state). 而 ér (and, but), 于 yú (at, in) and 之 zhī (a literary "of") rank 36th, 40th and 44th in writing and are missing from the spoken top hundred: they belong to a formal style that people seldom speak.

So a character list built from written text is the right list for reading, and the wrong one for following a conversation.

One character, more than one reading

Most characters have one pronunciation. A few of the commonest have two or three, and which one you need depends on the word:

得 is de in 觉得 juéde (to feel) and 记得 jìde (to remember), dé in 得到 dédào (to get), and děi when it means "must".
了 is le as a particle and liǎo in 了解 liǎojiě (to understand).
为 is wèi in 为什么 wèishénme (why) and wéi in 认为 rènwéi (to think).
行 is xíng in 不行 bùxíng (no way) and háng in 银行 yínháng (bank).
不 and 一 keep one reading but change their tone with the syllable after them: 不是 is said búshì, 一定 yídìng. The rules are on the Mandarin tones explained.

A character that only spells names

斯 sī is the 87th commonest character in dialogue, ahead of 快 (fast) and 死 (to die), although the CC-CEDICT entry for it reads only "(phonetic)" and "this". It is the standard character for writing an s sound in foreign names, and subtitled films are full of foreign names: 詹姆斯 Zhānmǔsī (James), 克里斯 Kèlǐsī (Chris). The written list puts it at 168. It is a reminder that this list comes from subtitles, many of them translated from other languages.

What to do with the list

Learn the characters inside words. The commonest words built from these hundred characters, 什么, 知道, 现在, 如果, 我们, are worth more than the characters one by one, and a character is often easier to remember once you know two words that use it.

Learn the particles early. 的, 了, 吗 and 吧 are all in the top 35 of speech, and 呢 is in the hundred. They carry no dictionary meaning you could translate word for word, and they turn up constantly.

Use a list made from speech if speaking is the goal. The dialogue list and the written list share most of their top hundred, but not the parts that make a conversation sound like one.

Verbavia (our own course) teaches Mandarin from a frequency list of words, each shown with its characters and its audio, with the grammar explained in English and reading, listening and writing sections in hand-written lessons. The Mandarin course is free on the web in any browser, including on a phone. You can browse the 1,000 most common Mandarin words, or take the two-minute Mandarin vocabulary test to see how many of them you already know.

How big the Chinese lexicon is in total is on how many words are in the Chinese language, and whether all this makes Mandarin hard is weighed up in is Mandarin hard to learn.

Caveats

Simplified characters only. The corpus is in simplified Chinese. A traditional-character corpus would count 们 as 們 and 这 as 這, and the order would differ a little.
Subtitles, not recorded speech. Subtitles are close to conversation, but many are translations of foreign films, which is why 斯 ranks so high, and film plots shape the counts: 死 sǐ (to die) is in the top hundred.
Swearing is counted too. After 他们 (they), the commonest longer word containing 他 is 他妈的 tāmāde, which CC-CEDICT glosses as a taboo curse, "damn it!" or "f*cking". It is in the table because it is in the dialogue.
Readings are a judgement where the uses split. For 得 and 地, the reading given is the one we judge commonest in speech, and the note gives the others.

Sources

Lison, P. & Tiedemann, J. (2016). OpenSubtitles2016: Extracting Large Parallel Corpora from Movie and TV Subtitles. Proceedings of LREC 2016. aclanthology.org
Dave, H. FrequencyWords: frequency lists from OpenSubtitles 2018 (zh_cn), content licensed CC BY-SA 4.0. github.com
Da, J. Modern Chinese character frequency list (现代汉语单字频率列表), data of 2004, from a large corpus of Chinese texts collected online. lingua.mtsu.edu
CC-CEDICT, the Chinese–English dictionary, MDBG release of October 2026 (125,244 entries), licensed CC BY-SA 4.0: every reading and meaning in the table was checked against it. mdbg.net
Wikipedia: Transcription into Chinese characters — 斯 sī as the standard character for s in foreign names.
Wikipedia: Chinese grammar — 们 as the plural marker for pronouns and nouns for people; 没 for negating 有 and completed events; 把 placing the object before the verb.
Wikipedia: Standard Chinese phonology — the tone changes of 不 and 一.
The counts, the "alone" share, the commonest words and the written-list ranks are computed by scripts/frequency_top.py in Verbavia's repository and printed from its output rather than typed.

Learn Mandarin Chinese

Start the free Mandarin Chinese course

9 languages, each taught from a frequency list, with the grammar explained properly and reading and listening built from words you already know. One lesson a day, free. See all 9.

The 100 most common Chinese characters

How the characters were counted

The 100 most common Chinese characters

Characters that are almost never words

Spoken Chinese is not written Chinese

One character, more than one reading

A character that only spells names

What to do with the list

Caveats

Sources

Learn Mandarin Chinese

More from the blog