Skip to main content

Home › Chinese Transliteration

Chinese to English Transliteration

Chinese transliteration means writing Chinese characters with the Latin alphabet according to how they are pronounced. The standard system for Mandarin is Hanyu Pinyin, and it is what this tool produces. Pinyin is not a translation: it tells you how a word sounds, not what it means.

0 characters
Result will appear here...

Successful conversions aren't stored. How this works

Fast Free No registration required

"Chinese to English" usually means Pinyin

This is the most common thing people are looking for, and the phrasing is a little misleading. Converting 中国银行 gives you:

中国银行 zhōng guó yín háng

zhōng guó yín háng is not English. It is the Mandarin pronunciation written in letters an English reader can attempt. The translation would be Bank of China; the transliteration is zhōng guó yín háng. If you need the meaning rather than the sound, what you want is a translator, not this tool.

Why character-by-character conversion is not enough

Many Chinese characters have more than one reading, and which one is correct depends on the word the character appears in, not on the character itself. 行 is a clear case: in 银行 (bank) it is háng, and almost everywhere else it is xíng.

Characters whose reading changes depending on the word
Character One word Another word
行 银行 → yín háng (bank) 他很行 → tā hěn xíng (he's capable)
乐 音乐 → yīn yuè (music) 快乐 → kuài lè (happy)
长 长城 → cháng chéng (Great Wall) 长大 → zhǎng dà (grow up)
重 重庆 → chóng qìng (Chongqing) 重要 → zhòng yào (important)

Getting these right needs a dictionary of words, not a table of characters. This is why the tool uses overtrue/pinyin, which matches whole words first, rather than the generic Unicode transliteration we use for other scripts. We tested the generic route and it returns "yín xíng" for 银行, which is wrong.

Chinese names are a separate case

Some characters have one reading in ordinary text and a different one when they are a surname, and the surname reading is the rarer of the two. Converting a name with the normal rules gives a plausible-looking result that a Chinese reader would recognise as wrong.

Surname readings compared with ordinary text readings
Name Read as a name Read as ordinary text
单田芳shàn tián fāngdān tián fāng
仇小山qiú xiǎo shānchóu xiǎo shān

Tick "This is a personal name" under the input box above and the tool switches to surname readings. Leave it off for anything that is not a name, because then it applies surname readings where they do not belong.

Pinyin, Wade-Giles and Bopomofo

Pinyin is the official romanization in mainland China and the international standard, but it is not the only system you will come across. Older English spellings of Chinese words usually come from Wade-Giles, which is why some familiar words look nothing like their Pinyin:

The same words in Pinyin and in Wade-Giles
ChinesePinyinWade-Giles
太极tài jít'ai-chi
道dàotao
功夫gōng fukung-fu
毛泽东Máo ZédōngMao Tse-tung

Bopomofo (Zhuyin Fuhao) is the third system you may see, taught in Taiwan and written with its own symbols: ㄅ ㄆ ㄇ ㄈ. It records the same sounds, but since it does not use Latin letters it is not a romanization, so it falls outside what this tool does.

What the tone marks are for

Mandarin has four tones plus a neutral one, and the tone is part of the word rather than decoration on it. The same syllable with different tones is different words:

The syllable ma in each of the four tones
PinyinCharacterMeaning
mā妈mother
má麻hemp
mǎ马horse
mà骂to scold

This is why the tool keeps the diacritics instead of stripping them. Pinyin without tones is still readable, but it is ambiguous in a way the original Chinese is not.

What this tool does not do

  • It does not translate. You get pronunciation, not meaning.
  • It does not go the other way. Turning Pinyin back into characters is a different problem: one Pinyin syllable matches dozens of characters, so choosing between them needs a predictive input method rather than a conversion rule.
  • It is not infallible on readings. Word-level matching resolves the common cases, but no automatic system gets every character right in every context.
  • It covers Mandarin. Cantonese and other varieties read the same characters differently and use their own romanizations, such as Jyutping.

Frequently Asked Questions

Writing Chinese characters with the Latin alphabet according to their pronunciation. In practice it almost always means Hanyu Pinyin, the official romanization of Mandarin.

No. Pinyin uses the same letters as English but spells out Mandarin sounds, so the result is not an English word and has no meaning in English. 中国银行 becomes "zhōng guó yín háng", which is how it is said, not what it means.

Because many characters have more than one reading, and the correct one depends on the word. 行 is "háng" in 银行 (bank) and "xíng" almost everywhere else. The tool matches whole words before falling back to individual characters.

Select Chinese and tick "This is a personal name". Several characters have a separate surname reading: 单 is "dān" in ordinary text but "shàn" as a surname, and without that option you would get the wrong one.

They are two romanization systems for the same sounds. Wade-Giles is older and is where English spellings like "t'ai-chi", "tao" and "Mao Tse-tung" come from; Pinyin writes those as "tài jí", "dào" and "Máo Zédōng". This tool uses Pinyin.

Not here. One Pinyin syllable corresponds to dozens of characters, so picking the right one requires a predictive input method rather than a conversion rule, which is a different kind of tool.

Last updated: 22 October 2026 · See what's changed

Other scripts

The same tool romanizes twelve other writing systems, and there is a longer explanation of what transliteration is if you want the general picture.

All languages What is transliteration?