How to use
The tool has two panes. The top pane takes a **Western date** (year/month/day) and returns the corresponding era and year-of-era — for example, 2024-04-01 → Reiwa 6 (令和6年), and 1989-01-07 → Showa 64 (昭和64年, the era's final day) but 1989-01-08 → Heisei 1 (平成元年). The bottom pane reverses: pick an era from the dropdown, enter the year-of-era, and the tool returns the corresponding Gregorian year along with the start date if you picked year 1 (元年).
Era transitions in modern Japan have happened on a specific calendar date — the day of imperial succession or proclamation — not on January 1st. This matters because **year 1 of any era is almost always a partial year**: Heisei 元年 ran from 1989-01-08 to 1989-12-31, only 358 days long; Reiwa 元年 began 2019-05-01 mid-year. The label "元年" (lit. "origin year") is the Japanese convention for year 1 — official documents and most newspapers write 令和元年, never 令和1年. Pre-Meiji eras (Edo, earlier) are not supported here; they used the lunisolar calendar with frequent era changes and complex calendar reform history. For dates before 1868-10-23, you need a specialized historical calendar tool.
Examples
Today's date in current era
Input
Western date: 2026-05-17
Output
era: 令和 (Reiwa, abbreviation R)
year-of-era: 8
formal: 令和8年5月17日
short: R8.5.17 or R08.05.17 (used on driver's licenses, residence cards)
Reiwa began 2019-05-01, so year-of-era = (current Western year) − 2018. The shorthand `R8` corresponds to 令和8年 and follows the convention on official Japanese documents — driver's licenses, juuminhyou (resident records), and tax forms use this format. Mixing the era abbreviation with the year (`R8`, `H30`, `S55`) is common in databases and admin systems. A common mistake in software: assuming year-of-era = Western_year − era_start_year + 1 without checking the start date — December 2018 is still Heisei 30, not Reiwa 0.
Showa → Heisei transition day-by-day
Input
date A: 1989-01-07 (final day of Showa)
date B: 1989-01-08 (first day of Heisei)
date C: 1989-12-31 (final day of Heisei 元年)
Output
date A: 昭和64年1月7日 (Showa 64, Jan 7) — Emperor Hirohito died this day
date B: 平成元年1月8日 (Heisei 元年, Jan 8) — Emperor Akihito succeeded
date C: 平成元年12月31日 (Heisei 元年, Dec 31)
Note: 昭和64年 only lasted 7 days. Heisei 元年 was 358 days.
The Showa → Heisei transition is the most-mishandled era boundary in Japanese software. Emperor Hirohito died on the morning of 1989-01-07; succession (and the new era name) was proclaimed the same day with effect from the following day. So 1989-01-07 is the only day of Showa 64, and the entire year 1989 is split into 7 days of Showa + 358 days of Heisei. Many naïve implementations bucket all of 1989 under Heisei or all of it under Showa, producing wrong outputs for ~ 2% of the year. The Reiwa transition (2019-04-30 → 2019-05-01) was less abrupt because Emperor Akihito abdicated rather than died, and the new era name had been announced one month in advance, but it still creates a mid-year boundary that bare year-math gets wrong.
Bulk Showa → Western conversion (work permit dates)
Input
era input: 昭和
years: S30, S45, S60, S64
use case: converting an older resume / employment history
Output
S30 = 昭和30年 = 1955
S45 = 昭和45年 = 1970
S60 = 昭和60年 = 1985
S64 = 昭和64年 = 1989 (Jan 1–7 only)
Formula: Western year = era_start_year + (year_of_era − 1)
Showa year = 1925 + (S − 1) = 1924 + S
Heisei year = 1988 + H
Reiwa year = 2018 + RJapanese resumes, employment records, and old certificates routinely use era years exclusively. The arithmetic shortcuts above work for the entire year *except* the transition year. Notable mnemonic: Showa is 1924+S, Heisei is 1988+H, Reiwa is 2018+R. When you convert in the other direction (Western → era), check that the resulting year is within the era's range — Showa 65 (1990) is invalid; that year is Heisei 2. Some Japanese government APIs (e.g. e-Stat) still return data with era-year fields; production code should validate the year falls within the era's active range or fail loudly rather than producing silently wrong dates.
FAQ
How are new era names chosen?
The 1979 **元号法** (Era Name Law) gives the cabinet the authority to declare a new era upon imperial succession. The procedure: a panel of scholars proposes candidates, the cabinet narrows the list, and the prime minister announces the final choice. Five criteria from the 1979 selection process are still used: (1) two kanji characters, (2) easy to write, (3) easy to read, (4) not previously used as an era name, and (5) reflects Japanese ideals. Reiwa (令和) was announced on 2019-04-01 with effect 2019-05-01, drawn from the Man'yoshu poetry anthology — the first era name drawn from a Japanese rather than Chinese classical source. The character 令 originally meant "auspicious" in this context, not the more common modern meaning "order/command", which caused some initial confusion in international media.
Why does year 1 of an era use 元年 instead of 1年?
It is a centuries-old Chinese and Japanese convention: the first year of any new reign is the **origin year** (元年), with subsequent years counted from there as 2年, 3年, and so on. The same convention persists in Korean royal histories (元年 / 원년) and Vietnamese imperial records. Official Japanese documents and most news media follow this rule consistently — you write 令和元年, 平成元年, 昭和元年, never 令和1年. Computer systems, on the other hand, almost universally store the integer 1 and only render it as 元 on display. The Microsoft Windows Japanese locale, the JIS X 0301 date standard, and the Unicode 元 character (U+5143) all support this convention but require careful handling: a date stored as "令和1年5月1日" and "令和元年5月1日" are semantically identical but lexically different.
How do Japanese government databases store dates internally?
Modern systems are migrating to Western dates internally and translating to era format only at the presentation layer — this avoids the 元年 / Western-year ambiguity at the database level. Legacy COBOL-era systems (still common in city halls, banks, and pension agencies) often store dates as a 7-digit packed BCD: era code (1–4 for Meiji–Heisei) + 2-digit year + 2-digit month + 2-digit day. The Reiwa era change in 2019 forced an industry-wide audit because many of these systems had hard-coded era code 4 = Heisei without a path for "5 = Reiwa" — the Japanese tech press extensively covered the multi-year remediation effort that preceded the transition. JIS X 0301:2002 specifies a string format `R01.05.01` for Reiwa year 1 May 1 (and earlier eras), still widely used in printed forms and many enterprise systems.
Do other Asian countries use a similar era system?
Japan is currently the only nation that *legally* uses an era system in everyday government documents. **South Korea** abolished gengo-style eras in 1962, switching to the Western calendar (단기 / Dangi years, counted from the legendary founding of Gojoseon in 2333 BCE, were also dropped except in some cultural contexts). **Taiwan (ROC)** uses 民國 (Minguo) years counted from 1912 — 2024 is Minguo 113 — used on driver's licenses and tax forms. **North Korea** uses 主體 (Juche) years from 1912 in official documents — 2024 is Juche 113. **Thailand** uses the **Buddhist Era** (B.E.) = Western + 543, so 2024 is BE 2567 — printed on every government form and most news sources. China abolished imperial era names in 1912; pre-1912 Chinese history is still cited by era (e.g. 康熙 reign 1662–1722), but no modern Chinese government uses them.
What about Emperor Akihito's posthumous era name?
When Emperor Akihito **abdicated** rather than died in 2019, he became 上皇 (Daijo Tenno, "retired emperor") and the Heisei era ended on 2019-04-30. Unlike his father Hirohito, who is now referred to as **昭和天皇** (Showa Tenno, posthumously named after his era), Akihito does not yet have a posthumous era-based name — he is referred to as 上皇明仁 or simply 上皇 while living. Upon his eventual death, convention is to refer to him as **平成天皇** (Heisei Tenno), aligning the era name with the imperial posthumous title. The same will eventually happen for Emperor Naruhito as 令和天皇 after his reign ends. This convention means the era name retroactively becomes the emperor's historical identifier — Showa Tenno, Heisei Tenno — though that final naming is only formalized after death, not at the era transition.
How do I implement Japanese era support in code?
For modern languages, use a library that already handles the transition dates correctly: **JavaScript** — the `Intl.DateTimeFormat("ja-JP-u-ca-japanese", {...})` Intl API returns era-formatted dates natively. **Python** — the `datetime` standard library does not support eras; use `dateutil` or `python-japanese-calendar`. **Java** — `java.time.chrono.JapaneseDate` is built in. **Go** — use `golang.org/x/text/calendar/japanese`. **Ruby on Rails** — Rails 6+ ships with `wareki` formatting helpers. For systems without library support, hard-code the five era start dates (1868-10-23, 1912-07-30, 1926-12-25, 1989-01-08, 2019-05-01) and pick the era whose start ≤ target_date — that is exactly what this tool does. Test cases must include the boundary days: era transition days, December 31 of an era's last year, and year-1 (元年) handling.
Related concepts
Japan's era system descends from the **Chinese imperial era name (年號 niánhào) tradition**, which began in 140 BCE under Emperor Wu of Han with the era 建元. China changed era names frequently — often multiple times per emperor's reign in response to good omens, disasters, or political resets — generating thousands of distinct era names across two millennia. Japan adopted the system in 645 CE with the era 大化 and followed the Chinese practice of multiple-eras-per-reign until the **Meiji Restoration in 1868**, when the convention shifted to **one era per emperor** (一世一元). Since then, every emperor has had exactly one era: Meiji (1868–1912), Taisho (1912–1926), Showa (1926–1989), Heisei (1989–2019), Reiwa (2019–present). Korea and Vietnam followed similar trajectories, abolishing era names in the 20th century. Japan's persistence with the system is partly cultural — eras serve as generational markers in everyday speech ("Showa-style" connotes mid-20th-century aesthetics) — and partly administrative inertia: government IT systems and printed forms baked in era support across the 1970s–1990s computerization wave.
The **計算機の Y2K-style problem** for Japanese eras came in 2019 with the Reiwa transition. Many enterprise systems had hard-coded the Heisei start year (1989) or the era count (4 modern eras) into business logic; adding a fifth era required code changes throughout banking, payroll, pension, and government systems. Microsoft published guidance and Windows updates specifically for the Reiwa transition; the Japanese government provided official era data files for use by software vendors. The transition was unusually well-prepared because the new era name was announced **one month in advance** of the calendar transition, the first time in Japanese history that the next era name was known before the previous era ended. Future transitions, occurring on imperial succession, may not have the same lead time if a reigning emperor dies unexpectedly — the contingency this tool inherently handles by simply consulting the era start dates.
Three adjacent **calendar systems** intersect with Japanese era usage. The **Western Gregorian calendar** is the primary internal calendar of all Japanese government systems since the 1873 Meiji adoption of Western dating; era format is a display layer. The **Japanese fiscal year (会計年度)** runs April 1 to March 31, not January 1 to December 31, which adds another layer when era-year + fiscal-year both appear in financial statements (e.g., 令和5年度 = April 2023 to March 2024). The **traditional lunisolar calendar (旧暦)** still drives some festivals (Setsubun, Higan, traditional new year in older communities) and Japanese astrology — it has 12 months that drift relative to the solar year, requiring occasional leap months. This tool covers only the post-Meiji era system; pre-Meiji eras like Edo (江戸時代, 1603–1868) use the lunisolar calendar and require entirely different conversion logic, including reform-era jumps like the 1872-12-02 to 1873-01-01 calendar reform that skipped most of December 1872.
