Romaji Converter

How to Use the Romaji Converter

1. 1. Pick what to convert. Kana → Romaji gives Hepburn. Romaji → Kana reads romaji and writes hiragana or katakana. Hiragana ↔ Katakana shows both scripts for whatever you paste.
2. 2. Paste your text. Put spaces between words, as romanized Japanese does — kana has none, and the tool cannot guess where words end. Kanji are left as they are and counted.
3. 3. Set the options. Choose how long vowels appear (macrons, no marks, or as spelled in kana) and whether a lone は or へ is read as the particle wa or e. Going back to kana, choose the script and the spelling system.
4. 4. Copy the result, or check it against the kana chart below it.

Which Romanization Is This?

The output is revised (modified) Hepburn, the system the Library of Congress uses in its ALA-LC Japanese table (2022 version), which in turn follows the American standard ANSI Z39.11. It is the usual choice for English-language writing and for catalogs. Two other systems turn up, and the tool reads their spellings when going back to kana: the older traditional Hepburn, which writes ん as m before b, m and p (shimbun), and Kunrei-shiki (ISO 3602, the Japanese government's system), which spells syllables by their place in the kana table.

The Rules the Tool Applies

• Long vowels. In the Library of Congress table, ああ, うう, ええ, おお and おう take macrons (okāsan, rōdō, onēsan, ōkami), and any vowel before a lengthening bar ー does too (konpyūta). Long i is the exception: いい stays ii (Niigata), but ビード is bīdo.
• ん. Always n, never m. Before a vowel or y it takes an apostrophe so the n does not join the next syllable: hon'yaku, shin'etsu.
• Small っ. Doubles the next consonant: kippu, zasshi. Before ち it adds a t instead (matcha), and ts and sh double only their first letter (mittsu, issho).
• Particles and old kana. A lone は is wa and a lone へ is e when the option is on. を is o, ゐ is i and ゑ is e.
• Extended katakana. Sounds for foreign words, such as ファ fa, ティ ti, ヴァ va and チェ che, follow Table B of the LC chart.

What No Converter Can Do

Kana records sounds, not meaning, and Hepburn records sounds, not spelling — so some choices belong to the reader. The converter cannot tell where one word ends and the next begins, which matters because the LC table writes vowels across a word boundary separately (koushi, not kōshi, for 子牛). It cannot see that the う in 思う is a verb ending rather than a long vowel, and it cannot read kanji at all. Type the kana reading for a kanji word, add the spaces, and correct a long vowel by hand when you know the word's parts.

• Romaji to kana is a best guess. ō might be おう or おお; the tool picks おう, the commoner, and you can switch it. Lone wa, e and o become わ, え and お, not the particles は, へ and を.
• Macrons in katakana become ー. kōhī is コーヒー, as in everyday writing.
• Kunrei or Hepburn? ti is ティ in the Hepburn setting and ち in the Kunrei one.

Hiragana and Katakana

The two scripts have the same sounds, so converting between them is a one-to-one swap. Hiragana is used for grammar and native words, katakana for foreign words, names and emphasis. The tool turns half-width katakana (ﾊﾟｿｺﾝ) into normal width first, keeps the lengthening bar ー, and writes ヴ as ゔ going the other way. Katakana-only letters such as ヷ have no hiragana and are left alone.

Japanese Titles in Citations

A reference to a Japanese book or article usually gives the title in the original script plus a romanized version, and often an English translation. Library records use ALA-LC, so romanizing with the same table makes your entry match the catalog. How to Cite a Source in Another Language shows where each part goes in APA, MLA and Chicago, and the Book Citation Generator formats the finished entry.

Frequently Asked Questions

What is the difference between romaji and Hepburn?

Romaji just means Japanese written in Latin letters. Hepburn is one system for doing it: it spells sounds the way an English reader would expect (shi, chi, tsu, fu). Others, like Kunrei-shiki, spell by the kana table instead (si, ti, tu, hu).

How do I type long vowels like ō and ū?

Copy them from the converter's output, use your keyboard's macron layout, or add them from your system's character map. If macrons are not allowed, choose the no-marks option for Tokyo, or the as-spelled option for Toukyou.

Why is there an apostrophe in 'hon'yaku'?

ん followed by a vowel or y is written n' so that you read hon-yaku and not ho-nya-ku. Without it, ほんや (hon'ya) and ほにゃ (honya) would look the same.

Why is は written ha here and not wa?

Because inside a word it is ha, as in はな. Only the grammatical particle is wa. The tool treats a は that stands alone between spaces as the particle; when it sits in the middle of a run of kana it cannot know, so write your own spaces.

Can the converter read kanji?

No. A kanji can have several readings, and picking the right one needs a dictionary and the context. Type the kana reading instead. The tool counts any kanji it leaves behind.

Is ou the same as ō?

In sound, usually. Revised Hepburn writes おう as ō, so 東京 is Tōkyō, while the as-spelled style writes it Toukyou. The Library of Congress table makes an exception when the う starts a new word or verb ending (思う is omou), which a program cannot see.

Which is the correct spelling for a passport?

Japanese passports do not use macrons, and the Foreign Ministry permits, without requiring, oh for a long o (Satoh). The no-marks setting gives the plain form (Sato); for the oh spelling, edit the result by hand.