ELOT 743 is the official Greek standard for romanizing Greek into Latin letters, also adopted by the UN. It is used on Greek passports and road signs. It transliterates most letters directly, for example η→i, υ→y, and ου→ou.

How does ELOT 743 differ from ISO 843?

ISO 843 is the international transliteration standard and uses diacritics to keep a reversible one-to-one mapping, for example ē for η and ō for ω. ELOT 743 favours plain Latin letters without diacritics, which is friendlier for everyday use and passports.

What does phonetic mode do?

Phonetic mode approximates how the Greek sounds in modern pronunciation. It renders context-sensitive groups such as μπ as b, ντ as d, γκ as g, and αυ/ευ as av/ev or af/ef depending on the following sound.

Why do some letters depend on context?

Greek has combinations whose sound changes with neighbours: γγ and γκ are 'ng/g', αυ is 'av' before voiced sounds but 'af' before voiceless ones, and final or initial position can matter. The tool applies these rules so the output reads correctly.

Is the conversion exact?

ELOT 743 and ISO 843 are deterministic standards and the tool follows their letter tables closely. Phonetic transcription is an approximation, since real pronunciation has subtleties no single mapping captures perfectly.

Why does αυ sometimes become 'av' and sometimes 'af'?

The υ in αυ is a consonant sound whose voicing follows the next letter. It becomes 'v' before a voiced sound (vowels and β, γ, δ, ζ, λ, μ, ν, ρ) and 'f' before a voiceless one (θ, κ, ξ, π, σ, τ, φ, χ, ψ) or at word end. So αυτό → afto but αύριο → avrio.

Which scheme should I use for a passport or airline booking?

ELOT 743. It is the standard used on Greek passports and avoids diacritics, so it passes through Latin-only forms and travel systems cleanly. ISO 843's accented characters (ē, ō) are often stripped or rejected by booking systems, so never use it for identity or travel data.

What is the Greek Romanization?

Romanize modern Greek into Latin script under ELOT 743 (the official Greek/UN standard), ISO 843 transliteration, or a phonetic transcription. Handles digraphs like ου, μπ, ντ and context rules for γγ, γκ, and αυ/ευ. It runs free in your browser on Gera Tools, with nothing uploaded.

Greek Romanization — Gera Tools

Name: Greek Romanization
Creator: Gera Tools
License: https://creativecommons.org/licenses/by/4.0/

Romanizing Greek means writing it in the Latin alphabet. There is no single “right” way — it depends on whether you need the official passport spelling (ELOT 743), a reversible scholarly transliteration (ISO 843), or something that reads the way the word sounds (phonetic). This tool offers all three.

Choosing the right scheme

The three schemes serve different audiences and purposes:

ELOT 743 is the Greek state standard and the one adopted by the United Nations. It appears on Greek identity documents, passports, and road signs. If you need a Greek name to match an identity document, ELOT 743 is the right choice. It avoids diacritics and unusual characters, so the output is compatible with most forms and databases.
ISO 843 is the international transliteration standard favoured by libraries, academic bibliographies, and archival work. It uses diacritics (ē for η, ō for ω) to keep the mapping reversible: given only the Latin output you can reconstruct the original Greek exactly. Choose it when the source language integrity needs to be preserved.
Phonetic mode is for readers who speak no Greek and want to approximate the pronunciation. It applies the context-sensitive rules that describe how words actually sound, not just their spelling — so μπ becomes b, ντ becomes d, and αυ/ευ flip between v and f depending on what follows.

How each romanization scheme maps letters

Each scheme is a mapping table from Greek letters and digraphs to Latin output, applied longest-match-first so multi-letter groups like ου, μπ, and αυ are handled before single letters.

ELOT 743 (official / UN): mostly plain Latin letters, e.g.

η → i    υ → y    χ → ch   ξ → x    ου → ou   αυ → av/af   ευ → ev/ef

ISO 843 (transliteration): reversible, using diacritics, e.g.

η → ē    ω → ō    χ → ch   ξ → x    υ → y

Phonetic: approximates pronunciation with context rules, e.g.

μπ → b    ντ → d    γκ → g    γγ → ng
αυ → av (before voiced) / af (before voiceless)
ευ → ev / ef    same rule

For αυ and ευ the tool inspects the following letter: before a voiceless consonant (θ, κ, ξ, π, σ, τ, φ, χ, ψ) the υ becomes f; otherwise it becomes v.

Comparison table for common Greek names

Greek	ELOT 743	ISO 843	Phonetic
Αθήνα	Athina	Athēna	Athina
Θεσσαλονίκη	Thessaloniki	Thessalonikē	Thessaloniki
Ευαγγελία	Evangelia	Euangelía	Evangelía
Αγγελική	Angeliki	Angelikē	Angeliki
Ξενοφών	Xenofon	Xenophōn	Xenofon

Notice that ELOT 743 and phonetic often give the same result for commonly used names, while ISO 843 diverges to show the full vowel distinctions (η vs ι vs υ, ω vs ο) that have merged in modern spoken Greek.

The two-consonant and vowel-cluster rules in detail

Greek romanization is hard precisely because several letter pairs behave differently from their parts. The rules the tool applies:

αυ / ευ / ηυ — the υ is a consonant sound here, not a vowel. It becomes v before a voiced sound (vowels and β, γ, δ, ζ, λ, μ, ν, ρ) and f before a voiceless one (θ, κ, ξ, π, σ, τ, φ, χ, ψ) or at word end. So αυτό → afto (τ is voiceless) but αύριο → avrio (ρ is voiced).
μπ — b at the start of a word (μπύρα → bira), often mp in the middle (αμπέλι → ampeli). Phonetic mode uses b/mp by position.
ντ — d initially, nt medially, by the same logic.
γκ / γγ — g or ng; γγ is always the ng sound (άγγελος → angelos), γκ is g initially and ng medially.
γ before ε/ι sounds — softens toward a y-glide in speech, which phonetic transcription approximates but the transliteration standards render as a plain g.

This is also where the schemes visibly disagree. ISO 843 is reversible: it never collapses η, ι, and υ into the same letter, because the whole point is to reconstruct the Greek. ELOT 743 and phonetic freely map all three to i, matching modern pronunciation but losing the distinction.

Which scheme to pick, by task

Your task	Use	Why
Match a Greek passport or ID	ELOT 743	It is the state/UN standard on documents
Fill a Latin-only web form or URL slug	ELOT 743	No diacritics, ASCII-safe
Library catalogue / academic bibliography	ISO 843	Reversible, preserves η/ι/υ, ω/ο
Help a non-Greek speaker pronounce a word	Phonetic	Renders μπ→b, αυ→av/af by sound
Reconstruct the original Greek later	ISO 843	Only reversible option here

A frequent real-world mistake is using ISO 843’s diacritic output (ē, ō) in a passport or airline booking, where the system strips or rejects the accented characters — always use the diacritic-free ELOT 743 form for identity and travel data.

Tips

ELOT 743 is the spelling that matches Greek passports and official signage, so use it for names that must match identity documents.
ISO 843 keeps a one-to-one mapping with diacritics, so it is the choice for library catalogues and bibliographies where the original spelling must be recoverable.
αυτό → phonetic afto because τ is voiceless; αύριο → avrio because ρ is voiced.
Greeklish (typing Greek words with Latin characters in informal messages) is different from any of these formal schemes and is not what this tool produces.

Sources and references

UNGEGN Working Group on Romanization Systems — Greek — the UN-adopted ELOT 743 romanization tables
ISO 843 — Conversion of Greek characters into Latin characters — the reversible transliteration standard with diacritics
ELOT (Hellenic Organization for Standardization) — the Greek national standards body behind ELOT 743, used on passports and road signs

Maintained by the Gera Tools editorial team. ELOT 743 and ISO 843 are deterministic and follow the letter tables above; phonetic mode approximates modern pronunciation (context rules for αυ/ευ, μπ, ντ, γκ) and is not a formal standard. Last reviewed 2026-07-02.