SayMei

Chinese vs Japanese: which is easier, and which should you learn?

We matched all 2,136 jōyō kanji to Chinese: 89% are written the same. Tones vs pitch accent, grammar, HSK vs JLPT, and which to learn for your goal.

Japanese borrowed its writing and many of its words from China. Its sounds and grammar stayed its own. That mix makes the choice more interesting than it looks.

This guide is for you if you're choosing between Mandarin and Japanese, or wondering how much one helps with the other. You'll get our own counts, the official numbers and a pick for each goal.

Key takeaways

  • The hard parts sit in different places. Mandarin is hardest for your ears and mouth; Japanese is hardest for grammar and reading.
  • Nearly nine in ten jōyō kanji are Chinese characters. We matched all 2,136 of Japan's everyday kanji, and most are written exactly as in Chinese.
  • Shared characters mostly keep their meaning. About one in twenty has drifted, and a few are true false friends.
  • A Japanese reader starts Chinese far ahead. Most characters up to HSK 4 are already familiar in some form, though every sound is new.
  • The top exams ask for similar amounts. JLPT's highest level once expected about 2,000 kanji, close to HSK 6.
  • Your goal should decide it. Japan, anime and Japanese companies point one way; China, Taiwan, Singapore and Chinese family point the other.

Most articles on this question give the same three facts: tones, three scripts, verb at the end. We wanted numbers, so we counted. Our data comes from Japan's official kanji list, Unicode's character database, China's exam syllabus and two open dictionaries. The method is in How we did this.

Is Chinese or Japanese harder to learn?

Neither is harder across the board. The difficulty sits in different places, and that's more useful to know than a single ranking.

The US State Department's language school puts both in its slowest group, Category IV. It budgets 88 weeks of full-time study for an English speaker to reach professional speaking and listening in either. The Defense Language Institute runs both as 64-week courses.

Here's where each language is harder for an English speaker, area by area:

  • Sounds

    Chinese harder

    Chinese
    Four tones on about 400 syllables
    Japanese
    About 100 sound units, with pitch accent
  • Grammar

    Japanese harder

    Chinese
    No verb endings; English word order
    Japanese
    Verb last, many endings, politeness levels
  • Scripts

    Japanese harder

    Chinese
    One, plus pinyin for learners
    Japanese
    Three: kanji, hiragana, katakana
  • Characters for everyday reading

    Chinese harder

    Chinese
    3,500 common characters
    Japanese
    2,136 everyday kanji
  • Readings per character

    Japanese harder

    Chinese
    Usually one
    Japanese
    About two on average
  • Speaking after 24 weeks

    Japanese harder

    Chinese
    1.5 on the government scale
    Japanese
    1.0
  • Full-time study to work level

    Even

    Chinese
    88 weeks
    Japanese
    88 weeks

Sources: State Department and DLI pages; Japan's 2010 jōyō list; China's 2013 character standard; Unihan; Chiswick & Miller 2004, reporting the State Department school's 1993 data. Our counts are explained under How we did this.

The speaking row is the surprise. After six months of full-time training, the school's Mandarin students spoke a little better than its Japanese students, in data from 1993 that Chiswick and Miller report. So Mandarin's tones didn't make it the slower language to speak, at least for those students.

Japanese wins on vocabulary, too. It has borrowed many English words and writes them in katakana: computer is コンピューター konpyūtā. Chinese more often translates the idea, so a computer is 电脑(diànnǎo), an "electric brain".

So the honest answer depends on the skill you care about first. If you want to chat soon, Japanese sounds come easier, but its grammar slows you down. If you want to read, Chinese asks for more characters, but each one usually has a single sound.

Nearly nine in ten kanji are Chinese characters

Japan's list of everyday characters, the jōyō kanji, has 2,136 entries. We checked each one against China's official list of Simplified characters and Taiwan's list of common Traditional ones.

A character counted as the same only if it's the same Unicode character. Unicode treats tiny regional differences in a stroke as one character, so 骨 "bone" counts as shared even though the printed shapes differ slightly. Shapes Unicode keeps apart, like Japanese 歩 and Chinese 步(bù) for "walk", count as different.

Here's how the 2,136 kanji compare:

Nearly nine in ten jōyō kanji are written exactly as in Chinese

Japan's 2,136 jōyō kanji, by how they compare with Simplified and Traditional Chinese characters

How the kanji comparesKanji
Same in Simplified and Traditional1,280
Same as Traditional only498
Same as Simplified only114
A Japanese form of a Chinese character232
No everyday Chinese counterpart12

Source: our count from Unicode's Unihan database 18.0, Japan's 2010 jōyō list and the KANJIDIC2 variant links. How we did this.

That's 1,892 kanji written exactly like a Chinese character, 89% of the list. Each group has a story:

  • Same everywhere: water, 水(shuǐ), mountain, 山(shān), and person, 人(rén), look the same in all three places.
  • Japan kept the old form: language, 語 in Japanese, is 语(yǔ) in Simplified Chinese, and car, 車, is 车(chē).
  • Japan simplified the same way: country, 国(guó), and study, 学(xué), are shared with mainland China but not with Taiwan.
  • Japan simplified its own way: air or spirit, 気 き ki, is 气(qì) in Simplified and 氣 in Traditional.
  • Made in Japan: the characters in to work, 働く はたらく hataraku, and mountain pass, 峠 とうげ tōge, were created in Japan.

If you read Traditional characters, Japanese looks more familiar than it does to a mainland reader. Taiwan's set matches 83% of the jōyō list; China's Simplified set matches 65%.

Most shared characters keep their meaning

Same form doesn't always mean same meaning. To check, we compared the English meanings two open dictionaries give for each shared character: KANJIDIC2 for Japanese, and CC-CEDICT for Chinese.

Six in seven share a meaning word outright. We read the rest by hand, and most just phrase the same idea differently. But 64 have drifted, and a random check of the matching group found a few more, like 私. It's "I" in Japanese, わたし watashi, and "private" in Chinese, sī.

So about one shared character in twenty has drifted. The rest keep a core meaning, which is why a Japanese reader can often guess a Chinese sign.

Whole words make the worst false friends

The bigger trap is words, where two familiar characters combine into something new. Here are some of the best known, checked in JMdict and CC-CEDICT:

In JapaneseIn Chinese
手紙 てがみ tegami, letter手纸(shǒuzhǐ), toilet paper
勉強 べんきょう benkyō, study勉强(miǎnqiǎng), reluctantly
大丈夫 だいじょうぶ daijōbu, all right大丈夫(dàzhàngfu), a real man
汽車 きしゃ kisha, train汽车(qìchē), car
愛人 あいじん aijin, lover爱人(àiren), spouse in mainland China
高校 こうこう kōkō, high school高校(gāoxiào), university
走る はしる hashiru, run走(zǒu), walk

Sources: JMdict and CC-CEDICT, downloaded October 2026.

Context usually rescues you in reading. In conversation it rarely does, because the shared words sound nothing alike.

A kanji reader starts Chinese far ahead

If you already read Japanese, Chinese reading comes fast. We took the 3,088 characters on China's new HSK syllabus and checked which ones a jōyō reader already knows.

"Same form" means the identical character. "In some form" adds the ones a reader knows in a Japanese or Traditional shape, like 语 for 語. Here's the share at each level, counting every character up to that level:

A kanji reader already knows four in five HSK 1–4 characters

Share of HSK reading-list characters, counted from HSK 1 up to each level, that match a jōyō kanji

Up toSame formIn some form
HSK 163.8%86.6%
HSK 260.4%85.7%
HSK 357.1%84.7%
HSK 454.4%82.5%
HSK 552.6%80.6%
HSK 651.2%77.5%
HSK 7–941.3%62.0%

Source: our count of the HSK syllabus's 3,088 characters to recognise against the jōyō list, via Unihan. "Up to HSK 4" means the 1,096 characters to recognise on the November 2025 syllabus; counts on a different HSK character list give slightly different shares. How we did this.

The gaps at HSK 1 are telling. The characters Japanese doesn't use are Chinese grammar words: you, 你(nǐ), the question word 吗(ma), the plural 们(men), this, 这(zhè), and very, 很(hěn). Japanese writes those ideas in kana.

One caveat: knowing the character isn't knowing the word. Every sound is new, and the tones are on top. "Library" is 図書館 としょかん toshokan in Japanese and 图书馆(túshūguǎn) in Chinese: you'd read it at once and not recognise it spoken.

It works the other way too. A Chinese reader opening a Japanese newspaper can follow its kanji-heavy headlines. The grammar between them, written in kana, is a separate job.

Japanese reads one character many ways

Chinese gives most characters one sound. Japanese often gives them two or more, and choosing the right one is a big part of learning kanji.

The jōyō list itself spells this out. It sets about two official readings per kanji, 4,388 in all, split between Chinese-derived on readings and native kun readings. In Chinese, we found that 87% of HSK characters have a single reading in China's standard dictionary.

The extreme case is life or birth, 生. In Chinese it's shēng. Japan's list gives it twelve readings, among them:

  • せい sei, as in 先生 せんせい sensei, teacher
  • しょう shō, as in 一生 いっしょう isshō, a lifetime
  • いきる ikiru, to live
  • うまれる umareru, to be born
  • なま nama, raw

Then come the kana. Japanese writes grammar endings and many native words in hiragana, and loanwords in katakana, with 46 basic signs each. They're far fewer than kanji, but a Japanese sentence mixes all three scripts, so you learn to read them together.

Chinese has one script. Learners lean on pinyin, the Latin spelling, but real Chinese text is characters only. So in Chinese, more characters to learn; in Japanese, fewer characters but more to know about each.

Tones are harder than pitch accent

Mandarin's sounds are the harder start for most English speakers. Every syllable has one of four tones, and the tone changes the word: 妈(mā) is "mum", 马(mǎ) is "horse", and 骂(mà) is "scold".

Japanese has pitch too, but it does less work. In Tokyo speech, each word has a fixed pitch pattern, and only some pairs depend on it:

  • chopsticks, 箸 はし hashi, starts high and falls
  • bridge, 橋 はし hashi, starts low and rises
  • rain, 雨 あめ ame, starts high and falls
  • candy, 飴 あめ ame, starts low and stays up

Source: Tokyo patterns as recorded by the NHK and Daijirin accent dictionaries, via Wiktionary. Kyoto and other regions differ.

Get Japanese pitch wrong and you'll sound foreign, but you're usually understood. Get a Mandarin tone wrong and you may say a different word. WALS, a database of language features, codes Mandarin's tones as complex and Japanese pitch as simple.

There's also the size of the sound system. Mandarin uses about 400 syllables, which our pinyin chart lays out, or about 1,300 once tones are counted. Japanese builds its words from about a hundred short sound units, called morae, and has five vowels.

One bonus if you already speak Japanese: Japanese listeners heard Mandarin's falling fourth tone better than English listeners did in one study. Tones are trainable either way, and our guide to Chinese tones shows how.

Chinese grammar is closer to English

Mandarin's word order mostly follows English, and its verbs never change form. Japanese puts the verb last and builds meaning onto its end.

Here's the same sentence, "I'm going to the library tomorrow":

  • Chinese: 我明天去图书馆。 Wǒ míngtiān qù túshūguǎn. Word by word: "I tomorrow go library."
  • Japanese: 明日、図書館に行きます。 あした、としょかんにいきます。 Ashita, toshokan ni ikimasu. Word by word: "Tomorrow, library to go."

Japanese verbs change form to show tense, negation and politeness. Chinese adds a separate word or says nothing:

MeaningJapaneseChinese
eat食べる たべる taberu吃(chī)
don't eat食べない たべない tabenai不吃(bù chī)
ate食べた たべた tabeta吃了(chī le)
eat, polite食べます たべます tabemasu吃(chī)

Sources: JMdict and CC-CEDICT. The Chinese 了(le) marks a finished action, not past tense as such.

WALS counts zero to one grammatical category on a Mandarin verb and four to five on a Japanese one. Japanese also has keigo, a system of polite and humble verb forms you'll need at work.

Chinese has its own hurdles: measure words, and the particle 了(le). Our guide to whether Chinese is hard covers them. Still, a Chinese beginner can say a lot before learning a single verb form.

The exams climb to similar heights

Each language has one main proficiency exam. China's is the HSK, which switches to a new nine-level form, HSK 3.0, this December. Japan's is the JLPT, with five levels from N5 up to N1.

The two bodies publish different things. The HSK syllabus lists every word and character. The JLPT stopped publishing lists when it was redesigned in 2010. Its last official counts are for the old levels, which it matches to today's N levels.

Here's how many characters each level expects you to read:

JLPT N1 expects about as many kanji as HSK 6 expects characters

Characters to read by each level, counted from the start. JLPT figures are the old levels' 'around' counts, matched to N levels by the JLPT.

LevelCharacters
HSK 1246
HSK 2371
HSK 3655
HSK 41,096
HSK 51,527
HSK 61,940
HSK 7–93,088
JLPT N5100
JLPT N4300
JLPT N21,000
JLPT N12,000

Sources: CLEC's HSK syllabus, November 2025; the JLPT's 2008 test guide, archived, and its FAQ. N3 has no official count.

Words tell a similar story: the old top JLPT level expected around 10,000 words, and HSK 7–9 lists 11,000. The JLPT's old guide also gave study hours for each level. The new HSK publishes none.

Don't read the chart as a conversion table. No official table links HSK levels to JLPT levels. The JLPT does now print a CEFR reference level on score reports, up to C1 for a high N1 score.

Japan's exam is the bigger one. About 1.9 million people applied for the JLPT in 2025, a record. China's testing body reports more than 810,000 HSK takers a year, China News Service wrote. For the HSK in detail, see our HSK exam guide.

Who speaks each language, and where

Mandarin has far more speakers; Japanese is concentrated in one rich country. Which matters more depends on where you'll use it.

In mainland China, nearly nine in ten people could use Putonghua, standard Mandarin, in 2025, according to official statistics reported by People's Daily. Mandarin is also the main language of Taiwan and one of Singapore's official languages. Ethnologue's widely quoted estimate is about 990 million native speakers.

Japanese is spoken almost entirely in Japan, home to just under 123 million people, by the Statistics Bureau's estimate. It travels well as a subject, though. In its latest survey, the Japan Foundation counted 4 million people learning Japanese at schools and other institutions abroad, a record.

So for numbers of people, Mandarin wins easily. For one market and its media, Japanese offers a lot from a single country.

Should you learn Chinese or Japanese?

Learn the one tied to something you'll actually do. Motivation keeps you going through years of study, and these two both take years.

Choose Japanese if

  • You're drawn to Japan. Anime, manga, games, travel or a job with a Japanese company give you a reason to study every day.
  • You'll live or work in Japan. Japanese is the language of daily life there, and the JLPT is the certificate employers know.
  • You want to speak early. The sounds are friendlier, and you can be understood without perfect pitch.
  • You already read Chinese. Most jōyō kanji will be familiar, so reading comes quickly.

Choose Chinese if

  • Your people speak it. Family in China, Taiwan or Singapore, a partner or colleagues decide this fast.
  • You want reach. Mandarin opens mainland China, Taiwan, Singapore and Chinese communities worldwide.
  • You'd rather keep grammar simple. No verb endings and English-like order make early sentences easier.
  • You already read Japanese. You'll recognise most beginner characters and can focus on sounds and tones.

If nothing ties you to either, try both for a week. Listen to each for ten minutes a day and notice which one you look forward to. That's a good sign of the one you'll keep up.

Can you learn both?

You can, but not at the same time as a beginner. Shared characters make it easy to mix up readings, and both need a lot of daily hours.

If you want both, get one to an intermediate level first. Then the second goes faster, because the characters and many borrowed words carry over. For how long each stage takes, see how long it takes to learn Mandarin.

Where SayMei fits, and where it doesn't

SayMei teaches Mandarin only. It doesn't teach Japanese, kanji readings or pitch accent, so if Japanese is your pick, look elsewhere.

For Japanese, the Japan Foundation's Irodori course and NHK's Easy Japanese are free, and the JLPT posts sample questions. We checked that these pages were up on 27 October 2026, but we haven't compared them in depth.

If you pick Mandarin, SayMei is built for the hard start this guide describes. You speak with Mei Lin, your AI teacher, from the first minute, and she listens to your tones. SayMei Premium costs $5.99 a month for 200 minutes of lessons, as our pricing page listed it in late October.

It won't replace a reading course. SayMei teaches Simplified characters and standard mainland Mandarin by default; for Traditional characters, see Simplified vs Traditional Chinese.

What to do next

  • Name your goal first. A country, a person or a kind of media should decide this, not difficulty.
  • Try both sounds for a week. Ten minutes of listening a day shows which one keeps you curious.
  • Pick one and stay with it. Get to an intermediate level before you start the other.
  • Use what you already read. Japanese readers can start Chinese with characters; Chinese readers can start Japanese with kana.
  • Start speaking early. Both languages reward saying words out loud from day one.

If Mandarin is your pick, see how many characters you'll need in how many Chinese characters do you need, or compare it with more languages in Chinese vs other languages.

FAQ

Is Japanese easier than Chinese for English speakers?

Officially they're equal: the US State Department budgets 88 weeks of full-time study for each. Day to day, Japanese sounds are easier, and its pitch accent rarely blocks understanding. Chinese grammar is easier, with no verb endings. Japanese reading is heavier per character, with about two readings each and three scripts. In the department's 1993 data, students spoke Mandarin slightly better than Japanese after six months.

Can I read Chinese if I know Japanese?

Partly. By our count on the new HSK syllabus's reading list, about 82% of the characters up to HSK 4 match a jōyō kanji in some form, and 62% of the full list do. You'll often guess signs and headlines. But Chinese grammar words like 吗(ma) and 们(men) aren't in Japanese, and the sounds are all different. Expect to read before you can understand speech.

Can Chinese speakers read Japanese?

They can follow kanji-heavy text, like headlines and signs, because 89% of jōyō kanji are written exactly as a Chinese character. Traditional readers have the edge: Taiwan's common set matches 83% of the list, against 65% for Simplified. But Japanese grammar lives in hiragana endings and particles, which a Chinese reader has to learn. Meaning can also mislead: 手紙 てがみ tegami is a letter, not toilet paper.

How many kanji are the same as Chinese characters?

By our count, 1,892 of Japan's 2,136 jōyō kanji, or 89%, are the same Unicode character as a Simplified or Traditional Chinese character. Most of the rest are Japanese forms of a Chinese character, like 気 き ki for 气(qì), and a dozen have no everyday Chinese counterpart. About one shared kanji in twenty has drifted in everyday meaning.

Should I learn Chinese or Japanese first?

Learn first the one you'll use soonest. If you plan to learn both, either order works, and the characters carry over both ways. Chinese first gives you a large character base and a simpler grammar to start with. Japanese first gives you an easier sound system and kanji that cover most beginner Chinese characters. What matters most is reaching an intermediate level before starting the second.

Is it a good idea to learn Chinese and Japanese at the same time?

For most beginners, no. The shared characters make it easy to swap readings, and each language needs regular hours to stick. We found no study comparing learning them together with learning one first. If you do both, separate them clearly, for example by time of day, and keep one as your main language.

How does HSK compare with JLPT?

The HSK publishes its word and character lists, which reach 11,000 words at the top. The JLPT stopped publishing lists in 2010, and its old top level expected around 10,000 words and 2,000 kanji. So the tops are close. No official table converts one exam to the other, but the JLPT now prints a CEFR reference level on its score reports.

Which is more useful for jobs, Chinese or Japanese?

It depends on your field and where you live. Mandarin reaches far more people, across mainland China, Taiwan and Singapore. Japanese is concentrated in one large economy, and the JLPT is the certificate Japanese employers ask for. We found no reliable worldwide data comparing job demand for the two. Look at job listings in your own field and city before you decide.

How we did this

Data, all open:

  • Jōyō kanji list, Agency for Cultural Affairs, 2010

    Licence
    Japanese government publication
    What we used it for
    The 2,136 kanji, their old forms and their 4,388 readings
  • Unihan 18.0

    Licence
    Unicode License
    What we used it for
    Jōyō, Simplified and Traditional sets, variant links, dictionary readings
  • HSK syllabus, CLEC, November 2025

    Licence
    Published by CLEC
    What we used it for
    The 3,088 characters to recognise, by level
  • KANJIDIC2 and JMdict, EDRDG, October 2026

    Licence
    CC BY-SA 4.0
    What we used it for
    Japanese meanings, readings and variants
  • CC-CEDICT, 8 October 2026

    Licence
    CC BY-SA 4.0
    What we used it for
    Chinese meanings and pinyin

Shared characters. We took the 2,136 code points Unihan marks as 2010 jōyō kanji, which match the official list. Simplified means China's 2013 Table of General Standard Chinese Characters, 8,105 characters. Traditional means Taiwan's 5,401 frequently used characters, the first level of the Big5 standard.

  • Same form: the identical Unicode character. Unicode merges small regional differences in a stroke, so those count as the same.
  • Japanese form: a kanji that matches neither set but links to a Chinese character through Unihan's variant fields, KANJIDIC2's variants or the jōyō list's own old forms.
  • No everyday counterpart: kanji made in Japan, such as 働, 峠 and 込, and a few used only in a Japanese sense.

We read every automatic match and corrected 11 wrong links by hand.

As a cross-check, we also counted Traditional as the Traditional forms of China's own list. That gives 1,857 matches instead of 1,778.

Meanings. For each shared character, we compared KANJIDIC2's English meanings with CC-CEDICT's glosses for the same single character and Unihan's definition. A pair matched if they shared a content word. Of the 1,892, 1,620 matched. We read the other 272 by hand and sorted them: 208 same core meaning, 45 shifted and 19 different. We also read a random 100 of the matches, and found three whose everyday sense differs. One person made these calls, so treat them as estimates.

HSK levels. For each HSK level, we counted the reading-list characters up to that level that are jōyō kanji, and those that are the Simplified form of a jōyō kanji. We used the November 2025 syllabus's list of characters to recognise. Counts on another HSK character list, such as a 1,200-character list for the first four levels, give slightly different shares.

Readings. Japanese: the readings column of the official jōyō list, 2,352 on and 2,036 kun readings. Chinese: the readings the 2013 General Standard Chinese Dictionary gives each HSK character, from Unihan.

Limitations.

  • No native speaker of either language has reviewed the examples yet.
  • The JLPT figures are the old levels' rounded counts from 2008, not an official list for today's test.
  • This guide reviews no apps and uses no SayMei learner data.

How to cite: SayMei (2026). Chinese vs Japanese: which is easier, and which should you learn? saymei.app/guides/chinese-vs-japanese.

Sources

The main primary sources, linked where they're used:

Methods

This guide uses the open datasets above and the official pages linked in each section, read on 27 October 2026. Every Chinese word was checked against CC-CEDICT and every Japanese word against JMdict, but neither has yet been reviewed by a native speaker. If you spot a mistake, tell us: we fix errors.

This guide reviews no apps and uses no SayMei learner data. We looked at the Japanese resources only by reading their public pages.