Jinmeiyō Kanji (人名用漢字): The 863 Name-Use Characters Beyond Jōyō
Jinmeiyō kanji (人名用漢字) are the 863 characters that Japanese law permits in personal names in addition to the 2,136 jōyō kanji. They are anchored in 戸籍法 (the Family Register Act) and administered as a Ministry of Justice ordinance.123 The list is the legal extension that makes name-traditional characters like 茜, 翔, and 凛 registrable at the koseki (family-register) window, even though they are not part of the general-literacy jōyō set.4
Overview
The 2,999-character name pool: jōyō plus jinmeiyō
Two government lists together define the legal name-character set: 常用漢字 (jōyō kanji) and 人名用漢字 (jinmeiyō kanji).14 The current jōyō total is 2,136 字種 (character types), set by 平成22年内閣告示第2号 (Cabinet Notice No. 2 of 30 November 2010).5 The current jinmeiyō total is 863 characters, fixed by the 25 September 2017 ministerial ordinance.167
戸籍法施行規則第60条 (Article 60 of the Family Register Act Enforcement Regulations) lists three categories of characters that may appear in a child's given name: characters in the 常用漢字表, characters in 別表第二「漢字の表」, and katakana and hiragana (excluding historical variants).34
MOJ-aligned reference works state the arithmetic plainly: 2,136 + 863 = 2,999 distinct kanji legal for koseki registration, alongside kana and the approved iteration marks (々, ゝ, ゞ, ー).46
What "jinmeiyō" means and what it does not
人名用 literally means "for use in personal names".46 The list gives legal permission only. 戸籍法 Article 50(1) requires that 「子の名には常用平易な文字を用いなければならない」 ("for the given name of a child, characters that are simple and in common use must be used"), and Article 50(2) delegates the scope to a Ministry of Justice ordinance.24
The Ministry of Justice ordinance is in 戸籍法施行規則 別表第二 (Appended Table 2 of the enforcement regulations). The koseki window checks each submitted given-name character against jōyō and jinmeiyō, and refuses anything outside both.13
The list states permissibility, not appropriateness. Many jōyō characters (子, 太, 美, 一) are far more common in Japanese given names than most jinmeiyō characters; the list is an extension, not a recommendation.46
Why a separate list exists at all
常用漢字表 is set by 内閣告示 (Cabinet notice) on an Agency for Cultural Affairs (文化庁) recommendation. It serves as a script-use guideline for laws, official documents, newspapers, magazines, and broadcasting.85 人名用漢字 is set by 法務省令 (Ministry of Justice ordinance) under the family-register administrative line, not the script-standard line. The two lists include many of the same characters, but they come from different agencies and different policy logic.48
The carve-out exists because 常用漢字 is calibrated for general literacy and public documents. Many name-traditional characters (渚, 萌, 凛, 茜) are not general-literacy candidates, but they have strong name tradition.49
The Family Register Law framing
戸籍法 第50条: the "constantly used and easy-to-understand" clause
戸籍法第50条 has two paragraphs. Paragraph 1: 「子の名には常用平易な文字を用いなければならない」 ("for the given name of a child, characters that are simple and in common use must be used").24 Paragraph 2: 「常用平易な文字の範囲は，法務省令でこれを定める」 ("the scope of characters that are simple and in common use is defined by Ministry of Justice Ordinance").24
The statute sets the standard (常用平易) but does not enumerate the characters; that enumeration is delegated to the ordinance.24 The official Ministry of Justice English translation of 常用平易な文字 is "characters that are simple and in common use".2
戸籍法施行規則 別表第二: where the list actually lives
戸籍法施行規則第60条 lists the three permissible categories. Sub-clause 2 cross-references 別表第二「漢字の表」, which contains the jinmeiyō characters.134 The MOJ official page 「子の名に使える漢字」 explicitly cites 「戸籍法施行規則第60条別表第二『漢字の表』」 as the location of the list.1
Updates to 別表第二 happen by 法務省令 (ministerial ordinance), not by Diet legislation.49 That procedure lets the list be amended one character at a time after a single family-court ruling.
Why the list is administered by the Ministry of Justice, not MEXT or the ACA
常用漢字 sits with the Cabinet, advised by 文化庁 (the Agency for Cultural Affairs, under MEXT), as a script-use guideline.5 人名用漢字 sits with 法務省民事局 (the Ministry of Justice Civil Affairs Bureau) because the family register (戸籍) is administered under the Family Register Act.148
The same character can therefore carry two policy histories: one as a literacy-calibrated script item (jōyō) and one as a name-register administrative item (jinmeiyō).48
What happens at the koseki window
A given name submitted on a 出生届 (birth-registration form) is checked by the municipal koseki office against the jōyō table and 別表第二.13 Characters outside both lists are refused. The family must either choose another character or file a 家事審判 (domestic-relations adjudication) petition with the family court. The 1997, 2015, and 2017 additions all arose this way.46101112
From 26 May 2025, the koseki office also checks the submitted reading (フリガナ) against the new "generally accepted reading" rule.131415
The expansion history: from 92 to 863 characters
The jinmeiyō list has been amended ten times since its 1951 debut. The timeline below tracks the running total of jinmeiyō characters at each step. The 2010 reshuffle is the only step that subtracts from the list. Characters moved into jōyō are still legal for names, but through jōyō rather than jinmeiyō. The May 2025 step is the only one that changes readings instead of characters.
1948: the original 92 characters under the new 戸籍法
The 1947–1948 revision of 戸籍法 (effective 1 January 1948) restricted given-name characters to those in 当用漢字 (1,850 字, 1946 内閣告示第32号).616 The first jinmeiyō list was issued on 25 May 1951 as an 内閣告示 (Cabinet notice). It added 92 characters on top of the 当用漢字 baseline.616
The dual structure (a literacy baseline plus a small name extension) has been in place ever since.46
1976: +28 → 120 characters
The 30 July 1976 ordinance added 28 characters, taking the running total to 120.616 It was the first significant post-1951 enlargement and reflected the period's parental pressure for wider name choice.6
1981: net +46 → 166 characters
On 1 October 1981, the literacy baseline grew from 当用漢字 (1,850) to the first 常用漢字 list (1,945 字, 内閣告示第1号).5 The 1981 ordinance moved 8 characters out of jinmeiyō because they were absorbed into the new jōyō list. It also added 54 new jinmeiyō characters, leaving a running total of 166.616
1990: +118 → 284 characters
The 1 April 1990 ordinance added 118 characters, taking the running total to 284.6 The expansion reflected the late-1980s wave of parental requests for more variety in name characters.6
1997: the 琉 ruling (Sapporo Family Court)
The 札幌家庭裁判所 (Sapporo Family Court) accepted 琉, an Okinawa-associated character then outside the list, as a legitimate name character following a parental petition.46 The 3 December 1997 ordinance added 琉, taking the running total to 285.6
The 琉 ruling was the first high-profile judicial signal that 常用平易 is a social-custom test, not a static-list test.4
2003 Supreme Court ruling: the 曽 case
最高裁判所第三小法廷 (Supreme Court Third Petty Bench), 平成15年12月25日決定, 平成15年(許)第37号 (the "曽良ちゃん命名事件" / Sora-chan naming case).1710
In November 2002, a father in Sapporo attempted to register his newborn son with the given name 曽良 (Sora), after Matsuo Bashō's disciple 河合曽良. 曽 was outside both 常用漢字 and 人名用漢字, so the koseki office refused the registration.10
The Court ruled that 常用平易な文字 must be judged by 社会通念 (social custom), not narrowly by the existing ministerial list. It also held that 曽 「は古くから用いられており、姓や地名にも多く使われ、国民の間に広く知られている字であって、明らかに常用平易な文字に当たる」 ("has been used since ancient times, is widely used in surnames and place names, and is widely known among the people, so clearly falls within the constantly-used and easy-to-understand characters").1710
The effect was structural: the ministerial ordinance was held invalid to the extent it excluded 曽. The MOJ added 曽 to 別表第二 on 23 February 2004, and the 488-character expansion followed in September.410
2004: +488 → 983 characters (the largest expansion)
The post-曽 sequence rolled out across the year: +1 (曽) on 23 February, +1 (獅) on 7 June, +3 (毘・瀧・駕) on 12 July, and then the 27 September 2004 法務省令 mass expansion of +488. It was the largest single addition ever.6109 Running total after September 2004 = 983.469
The 2004 expansion went through public consultation (deadline 9 July 2004). The initial pool of 578 characters was reduced after objections to characters such as 糞 (excrement), 屍 (corpse), 呪 (curse), 癌 (cancer), 姦, 淫, 怨, 痔, and 妾 (concubine). A further set of 79 characters (including 尻 and 嘘) was removed after deliberation on 13 August. The final ordinance on 27 September added 488 net.69
Representative 2004-era additions include 苺, 蕾, 絆, 琥, 珀, 凛, 萌, 渚, 茜, 翔, and 颯.467
2009: +2 → 985, then the 2010 reshuffle → 861
The 30 April 2009 ordinance added 2 characters (穹, 祷), taking the running total to 985.611
On 30 November 2010, the jōyō revision (平成22年内閣告示第2号) raised 常用漢字 from 1,945 to 2,136.5 The reshuffle pulled 129 characters from jinmeiyō into jōyō. They remain legal for names, now through jōyō inheritance. The reshuffle also pushed 5 jōyō characters into jinmeiyō.6 The running jinmeiyō total dropped from 985 to 861, even though the total name-eligible pool grew because jōyō gained more than jinmeiyō lost.6
The 2010 reshuffle is the reason raw jinmeiyō counts before and after that date are not directly comparable.6
2015: +1 (巫) and 2017: +1 (渾) → 863
The 7 January 2015 法務省令 added 巫.611 The addition followed a 家事審判: a couple in 松阪市 (Mie Prefecture) attempted to register a daughter's name containing 巫 in 2013. The 津家庭裁判所 松阪支部 (Tsu Family Court, Matsusaka Branch) ruled in their favour on 14 March 2014, holding that 巫 「は明らかに常用平易な漢字だ」 ("is clearly a constantly-used and easy-to-understand kanji"). The 名古屋高等裁判所 (Nagoya High Court) dismissed the city's immediate appeal in August 2014. The decision became final, and the MOJ amended the ordinance.11
The 25 September 2017 法務省令 added 渾.6127 Again the trigger was a 家事審判. Parents had been refused at the koseki window for a child's name containing 渾. The family court accepted the petition, the ruling became final, the MOJ ran a July–August 2017 public-comment period with no objections filed, and the ordinance was issued on 25 September 2017.12
The running total of 863 has been stable from 25 September 2017 onward.1127
2025: not a character expansion, but a yomikata reform
The amending statute is 令和５年法律第48号 (Act No. 48 of 2023), which amends 戸籍法. The amended provisions came into force on 26 May 2025.131415 The amendment introduces revised 戸籍法第13条, registering 氏の振り仮名 (family-name furigana) and 名の振り仮名 (given-name furigana) in the koseki for the first time.1315
It also introduces the substantive rule: 「氏名として用いられる文字の読み方として一般に認められているものでなければならない」 ("the reading must be one generally accepted as a reading for characters used in personal names").1315
The list of legal characters (the 863 + 2,136) is unchanged. The rule tightens the legal readings that may be attached to those characters at registration.1315 Examples of readings the MOJ guidance suggests would be refused include reading 高 as ヒクシ (a reading that means the opposite of the character), reading 太郎 as ジロウ or マイケル (a reading that names a different person), and reading 健 as けんいちろう (a reading that cannot plausibly be derived from the character).15
How court cases drive expansions
The 琉 case (Sapporo Family Court, 1997)
The 札幌家庭裁判所 decision established the judicial route in its modern form.46 Parents petition a 家庭裁判所 (family court) for a 家事審判 declaring that a specific character is 常用平易 despite being outside both lists. If granted, the koseki office must accept the registration. The MOJ typically follows by codifying the character in 別表第二.461112
The 曽 case (Supreme Court, 25 December 2003)
最高裁判所第三小法廷, 平成15年12月25日決定, 平成15年(許)第37号; the 「曽良ちゃん命名事件」.1710
The Court applied the 常用平易 test to social custom rather than to the ministerial list, and ruled that the ordinance was invalid to the extent it excluded 曽.1710
The doctrinal effect is that courts can compel the MOJ to recognise characters that are clearly in common social use, even before the ordinance is amended. This is the legal anchor for every post-2003 case-driven addition.17101112
Post-2004: case-by-case additions rather than mass expansions
After the 2004 mass expansion exhausted the obvious backlog of socially established characters, the route returned to single-character additions. 巫 (2015) and 渾 (2017) both arose from individual 家庭裁判所 petitions that the MOJ then codified by ordinance.61112
The pattern in plain terms: a parent files at the koseki window; the office refuses; the parent files a 家事審判 with the local family court; the court rules, and the appellate court affirms; the MOJ runs a brief public-comment period; the ordinance is amended.1112 The mass-expansion era ended in 2004. The case-by-case era began with 巫 in 2015.61112
What is not on the list and stays not on the list
Characters whose social-use test plainly fails were excluded from the 2004 expansion after public consultation: 糞 (excrement), 屍 (corpse), 呪 (curse), 癌 (cancer), 姦, 淫, 怨, 痔, 妾 (concubine), and around 79 further characters such as 尻 and 嘘 removed at a later deliberation stage.69
No court has since compelled the MOJ to add any of these. That exclusion is therefore stable: characters that are common in modern Japanese but socially incompatible with personal-name use.69
Name-reading conventions and the nanori bridge
Jinmeiyō characters often carry nanori readings the kanji's jōyō entry does not list
Standard 漢字辞典 (kanji dictionaries) split kanji readings into on'yomi, kun'yomi, and 名乗り (nanori). A name kanji can carry many nanori entries that do not appear in ordinary-vocabulary dictionaries.4 The J-Compass article on nanori readings covers the linguistic mechanics of this third reading layer in detail.
Examples of jinmeiyō characters with name-tradition readings beyond their dictionary on/kun include 颯 (on: ソウ / ハツ; nanori: ハヤテ); 翔 (on: ショウ; kun: かけ-る, とぶ; nanori: トブ, カケル); 凛 (on: リン; nanori: rare standalone-name usage); 茜 (on: セン; kun: あかね, frequently a standalone given name); and 苺 (kun: いちご, used as a girl's given name on the strength of the kun alone).47
颯 was added to jinmeiyō in the September 2004 expansion. ハヤテ is a nanori-style reading, not one of the character's dictionary on/kun entries.497
息子の名前は翔と書いて「カケル」と読みます。497
"Our son's name is written 翔 but read 'Kakeru'."
翔 entered jinmeiyō in the September 2004 expansion. Both ショウ (on'yomi) and カケル (a nanori-style reading) are accepted at registration under the legal-character and legal-reading framework.497
娘の名前は茜です。夕焼けの色から取りました。497
"Our daughter's name is Akane. We took it from the colour of the evening sky."
茜 was added in the September 2004 expansion. The reading あかね (kun) doubles as a free-standing given name, the most common pattern for nature-image jinmeiyō characters.497
The 2025 yomikata rule changes the reading layer, not the character layer
令和５年法律第48号 inserts the rule that the registered reading 「氏名として用いられる文字の読み方として一般に認められているもの」 must be a generally accepted reading.131415 The character set legal for names is unchanged. The readings legal for those characters are tightened.1315
This is the institutional response to the long tail of キラキラネーム (flashy or unconventional name) readings: cases where the chosen characters were jōyō- or jinmeiyō-legal, but the assigned reading was unmoored from the characters.1315
Why furigana on Japanese names is structural, not decorative
Because nanori readings frequently lie outside the dictionary on/kun entry, a written name in Japanese is not automatically pronounceable.4 This is why Japanese forms include a separate phonetic-reading field (フリガナ / ヨミガナ) for names. From 26 May 2025, that reading is also a koseki-registered datum.41315
The 2025 reform raises the furigana from a courtesy field to a legal one.131415
How jinmeiyō intersects the rest of the kanji landscape
2,136 + 863 = 2,999 name-eligible characters (plus kana)
The arithmetic is simple: 2,136 jōyō characters (current set since 30 November 2010) plus 863 jinmeiyō characters (current set since 25 September 2017) = 2,999 distinct kanji legal for given-name registration.145
戸籍法施行規則第60条 also permits hiragana and katakana for given names.3 The approved iteration marks for given names follow the same regulatory line; 々, ゝ, ゞ, and ー are accepted in koseki practice.34
What sits outside both lists: hyōgaiji
Hyōgaiji (表外字) means "characters outside the table": all kanji beyond 常用漢字 and 人名用漢字. This includes around 10,000+ characters in literary, classical, place-name, and surname usage.48 Hyōgaiji are not legal for new given-name registrations, but they appear freely in pre-existing family-name records in the koseki.48
Why family names (姓) are unrestricted but given names (名) are restricted
戸籍法第50条 applies to 子の名 (the given name of a child), not to 氏 (family name).24 Family names inherited from older koseki records (髙, 﨑, 邊, 邉) are legal because they predate current registration rules. Given names are constrained because each is a new entry tested against 別表第二.24
The variant-form (異体字) wrinkle
A subset of jinmeiyō entries in 別表第二 are variant forms (異体字) of jōyō characters: 桜/櫻, 国/國, 沢/澤, 高/髙 are recurring pairs.46
This is how a parent can register 國 or 櫻 in a given name, even though only the simplified form (国, 桜) is in 常用漢字.46 The 2010 jōyō revision absorbed many such standard forms into 常用漢字 while keeping the variants in 別表第二.65
Good to know
"Jinmeiyō" is a permission roster, not a popularity ranking
A frequent misreading treats inclusion on the jinmeiyō list as evidence that a kanji is popular in Japanese names. The list is a legal-permission roster. Most actual Japanese given names use jōyō characters (子, 太, 美, 一) far more often than they use jinmeiyō characters.46 The list is an extension to the literacy-calibrated jōyō baseline, not a popularity index.4
Pre-2010 and post-2010 jinmeiyō counts are not the same metric
The 2010 reshuffle moved 129 characters from jinmeiyō into jōyō. Those characters remain fully legal for given names. They are now reached via jōyō inheritance rather than jinmeiyō listing.6 The jinmeiyō headline number fell from 985 to 861 in 2010, but the total name-eligible pool grew because the new jōyō baseline gained more than the jinmeiyō list lost.65 Reporting that frames the 2010 step as a "reduction in legal name characters" gets the arithmetic backwards.
English-language press calls these "approved baby names", but the list approves characters
The MOJ list is a list of approved characters, not approved names. Parents combine characters and assign a reading. That combination is what becomes a name. English-language headlines often elide this distinction, especially in coverage of the 2025 reading reform.415
The case-by-case era is the only era after 2004
The mass-expansion era ended on 27 September 2004. Every addition since then (2009, 2015, 2017) has come from a 家事審判 petition followed by ministerial codification.61112 The Supreme Court's 25 December 2003 曽 decision is the legal anchor that makes this single-character route work. The post-2004 pattern is now stable.1710
"Kirakira names" are a reading problem, not a character problem
Names like 光宙 read as ピカチュウ used jōyō- and jinmeiyō-legal characters. The controversial part was the assigned reading.15 The 2025 yomikata (reading) rule (一般に認められている読み方) targets the reading layer that produced these cases without touching the character list.1315
Some jinmeiyō entries are variants (異体字) of jōyō characters
The list permits a variant form as a given-name character even when the standard form is in jōyō. 桜/櫻, 国/國, 沢/澤, and 高/髙 are recurring pairs.46 A parent who wants 國 or 櫻 in a given name registers the variant through jinmeiyō; the simplified form goes through jōyō.
Naming-rule conservatism varies by municipality
Within the legal framework, individual koseki offices have historically varied in how strictly they apply the 常用平易 test for borderline readings, especially before May 2025.415 The 2025 amendment partially standardises this by codifying the "generally accepted reading" test at statute level. The variance has narrowed, though not disappeared, since then.1315
See also
- Secondary School Jōyō Kanji (中学校 + 高等学校): The 1,110-Character Set Beyond Elementary
- How to Predict the Reading of an Unknown Kanji Compound: The On+On Default, Jūbako, Yutō, and the Look-It-Up Bucket
- Jukujikun (熟字訓): When a Compound's Reading Is Assigned to the Whole Word, Not the Kanji
- Go-on, Kan-on, Tō-on: The Historical Layers Behind a Kanji's Multiple On'yomi
- Ateji (当て字): Kanji Chosen for Sound, Not Meaning