International Morse Code Chart (ITU-R M.1677-1): Letters, Numbers, Punctuation, and Timing

The primary standard and its scope

The International Telecommunication Union's Radiocommunication Sector publishes Recommendation ITU-R M.1677. The ITU catalogue lists revision M.1677-1, approved in October 2009, as the in-force main recommendation at the time this guide was reviewed. Its stated scope covers International Morse characters and operational provisions for their use in radiocommunication services.

That scope matters. The recommendation is not a registry for every sequence ever used by a national alphabet, telegraph company, software package, amateur convention, or online puzzle. It prints the basic Latin letters, an accented E, decimal figures, selected punctuation, and several miscellaneous or procedural signals. This guide labels those categories instead of blending popular extensions into the same “official” table.

A written dot-and-dash chart is an abstraction of a timed signal. The standard prints dots and dashes for lookup, then defines their relative duration and the intervening silence. A text translator can preserve character and word boundaries with separators, but it does not reproduce sending speed, rhythm, tone frequency, channel conditions, or operating procedure.

International Morse code letters A–Z

Letter case is not encoded. The same group represents A and a, and a decoder commonly returns uppercase letters. Each code cell below is one character group; spaces shown inside this table separate rows or visual items, not signals within a character.

The recommendation labels ..-.. as accented E. Other language-specific letters seen in regional Morse charts require a separately named repertoire. A general translator should report an unsupported character rather than remove its accent, substitute a look-alike, or romanize it without permission.

International Morse code numbers 0–9

Every decimal digit uses five signals. From 1 through 5, dots accumulate from the left; from 6 through 0, dashes accumulate. That pattern helps with memorization but does not change the need for a boundary between adjacent digits.

For example, 2026 becomes ..--- ----- ..--- -.... when ordinary spaces divide the four digits. Without those spaces, the twenty-signal stream does not visibly preserve which five signals belong to each digit.

ITU punctuation and written signs

The table below follows the written punctuation and sign entries printed in Recommendation ITU-R M.1677-1. Some sequences have more than one textual name because the same sign served related telegraph uses. The multiplication sign uses the same sequence as X. A product that round-trips text therefore cannot infer from -..- alone whether the sender meant letter X or multiplication.

Miscellaneous signals in the ITU table

Several listed sequences represent an operating instruction rather than an ordinary printable character. A text translator should not force all of them into one-character output. Procedural signals can also be transmitted as a continuous group, so inserting normal letter gaps inside a displayed mnemonic would change the signal.

Duplicate patterns are unavoidable because meaning can come from context. -.- can be letter K or an invitation to transmit; -..- can be X or the multiplication sign. The LiveParse translator's ordinary-character repertoire excludes procedural signals. For -..-, the interface returns X while keeping an X/× ambiguity warning visible. That practical policy cannot recover an unstated intention from the signals alone.

Morse code timing and spacing

The relative unit rules are short enough to memorize and important enough not to approximate. If a dot lasts 100 milliseconds, a dash lasts 300 milliseconds. The silence inside a character lasts 100 milliseconds, between characters 300 milliseconds, and between words 700 milliseconds. Choose any consistent dot duration; multiply it by the ratios below.

Do not add the one-unit internal gap on top of a three-unit letter gap or add a three-unit letter gap on top of a seven-unit word gap. Each number describes the complete silence at that boundary. For A (.-), send one unit on, one off, then three on. If another letter follows, leave three units off before its first signal.

These ratios explain why Morse is not simply binary. Dot and dash are two signal lengths, but silence also has at least three meaningful lengths. A data representation can model that timing explicitly, or it can replace the two larger boundary gaps with text separators. Mapping dot to 0 and dash to 1 while dropping silence is not reversible.

Typed dots, dashes, spaces, and slashes

Plain text needs a convention because repeated ASCII spaces are not a calibrated clock. A practical format keeps the signals for one character adjacent, writes one ordinary space between characters, and uses / —a slash surrounded by spaces—between words:

.... . .-.. .-.. --- / .-- --- .-. .-.. -..

The slash here is metadata for a seven-unit word gap. It is not the transmitted slash punctuation character. To encode an actual slash between A and B, use the slash character's group as a normal character:

.- -..-. -...

A format may instead use multiple spaces, a newline, or another token for words. That is acceptable if it is documented and parsed unambiguously. Copy the exact convention expected by the receiving tool rather than assuming that every website treats whitespace identically.

How to translate text to Morse code

1. Preserve the source. Keep the original message so an unsupported character can be fixed without reconstructing it from partial output.
2. Choose a named repertoire. For this chart, use the characters and signals in ITU-R M.1677-1; agree separately on extensions.
3. Ignore letter case only. A and a share a code, but do not strip accents or transliterate other scripts unless that transformation is explicitly requested.
4. Map one character at a time. Add a visible character separator after each complete Morse group.
5. Preserve word boundaries. Emit a distinct word delimiter for each run of text whitespace according to the selected text format.
6. Stop or report unsupported input. Never silently remove a symbol, because the resulting message would say something different.
7. Review ambiguous printed signs. Context determines whether a shared pattern means X or multiplication, K or invitation to transmit, and so on.

For example, MORSE CODE becomes -- --- .-. ... . / -.-. --- -.. .. The mapping changes the representation, not the language or meaning. A French phrase remains French after its supported characters become Morse.

How to decode Morse code to text

1. Identify the notation. Confirm which characters mean dot and dash and which token separates words.
2. Split at word boundaries first. Do not let a delimiter slash become the five-signal slash punctuation code.
3. Split each word into character groups. Empty groups may indicate accidental repeated separators and should follow a documented policy.
4. Validate the symbols. Reject unexpected characters instead of deleting them and joining formerly separate groups.
5. Look up each complete group. Report the position of an unknown sequence so the source can be corrected.
6. Keep ambiguity visible. Where a pattern has procedural and printable meanings, do not claim context the signal did not provide.
7. Compare with the original timing when consequential. A transcription may have inserted the wrong boundary even when every resulting group is valid.

The group sequence .--. .-. .. ...- .- - . decodes deterministically as PRIVATE because each letter boundary is explicit. The concatenated stream .--..-......-..--. does not preserve that segmentation and can have other valid divisions. A dictionary can rank guesses, but probability is not proof of the sender's message.

Percent, quotation, minutes, and fractions

M.1677-1 also describes how to transmit written signs that lack a single corresponding signal. A percent sign is expressed through figures and the fraction bar, and the recommendation gives additional formatting for a preceding number. Quotation marks use their special signal before and after the quoted words; in a code-converter context the apostrophe can be repeated as specified by the recommendation.

Minute and second marks following figures use the apostrophe signal once or twice as appropriate. The recommendation expressly distinguishes that usage from the quotation-mark sequence. Mixed numbers and fractions also have prescribed hyphenation to preserve where the whole number ends and the fraction begins.

Those are operational transcription procedures, not one-character Unicode mappings. A simple converter should avoid claiming that every arbitrary symbol has a direct Morse equivalent. For formal traffic, read the relevant provisions in the official M.1677-1 document and follow the operating rules that govern the service.

What a written Morse translator cannot establish

A dot-and-dash translator does not listen to audio. It cannot separate a tone from noise, measure element duration, determine sending speed, or infer whether a pause was an intentional word boundary. Those tasks require signal-processing input and a model for timing variation.

It also does not authenticate a sender, conceal a message, or prove that a transcription is complete. Morse is a public code. If confidentiality or integrity matters, apply an established cryptographic protocol before choosing a transport representation.

Finally, a translator cannot supply operational judgment. Printed examples are suitable for learning, puzzles, documentation, and checking a transcript. Actual amateur, maritime, aviation, emergency, or other regulated communication must follow the relevant licensing, equipment, frequency, priority, and service rules.

A reliable Morse transcription checklist

1. Name the reference. Record ITU-R M.1677-1 or the exact alternative repertoire in use.
2. Keep raw timing when available. A text transcription loses duration evidence.
3. Mark two boundary levels. Distinguish character gaps from word gaps.
4. Separate delimiter and punctuation syntax. A slash used between words must not collide with encoded slash punctuation.
5. Report unknown groups. Do not discard them to create readable prose.
6. Report unsupported source characters. Do not silently remove, deaccent, or transliterate them.
7. Label extensions. Make non-ITU punctuation and national characters an explicit interoperability choice.
8. Preserve context. Shared patterns can mean a letter, written sign, or procedural signal.
9. Use real procedures for real traffic. A browser reference is not an operating authorization or emergency service.