How to decode Morse code by hand, step by step

To decode Morse code by hand, split the message into words at the slashes, split each word into characters at the spaces, and look up each character in a chart, writing the letter under its code as you go. Mark anything you cannot place, finish the rest, and come back to it with the whole word as context. A pencil, a sheet of paper and a chart sorted by length are all you need.

This guide is about Morse that is written down: a puzzle, a note, a bracelet, a line of dots and dashes in a book. Decoding by ear works differently, and there is a word on that at the end.

Step 1: Find the word breaks

Look for the slashes first. In the usual written form, / separates words, so a message with four slashes has five words. Some writers use a vertical bar, a wider gap or a new line instead; whatever the mark, find it and draw a clear line down the page between words.

If there are no word breaks at all, do not worry yet. Decode the letters in one long run and find the words afterwards, the way you would read a message with the spaces left out.

Step 2: Split each word into characters

Inside a word, a single space separates one character from the next. Copy the message onto your paper with extra room between the groups, one group per box if you like squared paper, and leave a line underneath for the letters.

Then count the marks in each group. The count is the fastest way into any chart, because Morse sorts neatly by length.

Step 3: Look up each character by its length

Every letter has one to four elements. Every figure has exactly five. Most punctuation marks have six. So once you have counted a group, you only have a handful of candidates:

Elements Characters
1 E ., T -
2 I .., A .-, N -., M --
3 S ..., U ..-, R .-., W .--, D -.., K -.-, G --., O ---
4 H ...., V ...-, F ..-., L .-.., P .--., J .---, B -..., X -..-, C -.-., Y -.--, Z --.., Q --.-
5 the figures 0 to 9; slash -..-.; plus or AR .-.-.; equals or BT -...-; open parenthesis or KN -.--.; ampersand .-...
6 period .-.-.-, comma --..--, question mark ..--.., apostrophe .----., hyphen -....-, close parenthesis -.--.-, quotation marks .-..-., colon ---..., at sign .--.-.

The ampersand is commonly used but is not in the ITU standard, and neither are the exclamation mark, -.-.--, and the semicolon, -.-.-., both six elements long.

In the letter rows, the codes are listed in order: every code that starts with a dit comes before every code that starts with a dah, and so on for the second mark and the third. So for a letter that starts with a dah, skip straight to the second half of its row. That is the idea behind the Morse tree, which lays the same order out as branches; the Morse code tree shows how to read it.

The figures need no chart at all once you see the pattern. For 1 to 5, the number of dits at the start is the figure: 3 is ...--. For 6 to 9, the number of dahs at the start is the figure minus five: 7 is --.... Zero is five dahs, -----.

Write each letter under its code as soon as you find it. Do not try to hold a word in your head; a pencil is faster than memory.

Step 4: Work through an example

Here is a short note, written the usual way:

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

Four slashes, so five words. The first two are quick: -- . . - is M, E, E, T, and -- . is M, E. So far: MEET ME.

The third word is one group, .--. Three elements, starting with a dit, then two dahs: W. The fourth word, - .... ., is T, H, E. So now the note reads MEET ME W THE, which is not English. Leave the W for a moment.

The fifth word has seven groups:

Group Elements Letter
..-. 4 F
--- 3 O
..--. 5 not in the chart
- 1 T
.- 2 A
.. 2 I
-. 2 N

The third group has five elements, so it should be a figure, but it is not one: 2 is ..---, and this ends in a dit. It is not one of the common punctuation marks either. Some languages add codes for accented letters, but in English the likeliest answer is a missing space: two letters have run together.

Try splitting it at each point. Three of the splits give a real letter on both sides:

Split Letters
. and .--. E P
.. and --. I G
..- and -. U N

FO, then EP, IG or UN, then TAIN. Only one makes a word: FOUNTAIN.

Now go back to the W. A lone W cannot be a word in this sentence, so try the same trick. .-- splits into .- and -, which is A and T. MEET ME AT THE FOUNTAIN.

Step 5: Fix run-together and split letters

The example shows the usual damage in written Morse: a space that went missing. The opposite happens too, a space where none belongs. A few rules catch most of it.

  • No letter has more than four elements. A group of five or more in the middle of a word of letters is either a figure, a punctuation mark, or two letters with the space lost. If it is not in the chart, it is most likely the last.
  • A figure among letters is a warning. The figure 5 is five dits, ....., which is exactly what HE, IS or SI look like with the gap missing. If a word decodes as T5, it is probably THE.
  • A word that is valid letter by letter but makes no sense often has one merged group. Short letters are the usual suspects: E and T are one mark each and disappear into their neighbors easily.
  • A letter that splits into two short ones is the opposite fault. If L, .-.., was written with a stray space, .- .., you get AI where an L belongs. When a word has an odd pair of short letters, try joining them.
  • Eight dits in a row, ........, is the error signal. The writer is saying “ignore that”. Drop the word before it and read on; the error signal explains how senders use it.
  • Some five-mark groups are prosigns. .-.-. at the end of a message is AR, end of message, not a plus sign. -...- between sentences is BT, a break.

When you cannot settle a group, write an underscore in its place and keep going. The rest of the message usually narrows it down.

Step 6: Check your answer

Read the decoded text aloud. Then encode a word or two back into Morse and compare it mark by mark with the original; a mismatch shows you exactly where a group was misread.

A decoder is a good second opinion, as long as you remember what it can and cannot do. It reads exactly what is written, so it will print W where the writer meant AT. Only you can see that the sentence needs fixing. On a Mac, Dot & Dash has a Transcribe screen where you type or paste dots and dashes into Morse to Text and the message appears as you go, so you can check each word against your pencil copy.

Decoding by ear is a different skill. There are no spaces to see, only silences to hear, and there is no time to count marks. If that is your goal, how to start copying Morse code by ear is the place to begin.

Questions

What is the fastest way to decode Morse code by hand?

Count the marks in each group and use a chart sorted by length, so you only compare a group with characters of the same size. With practice you will know the one, two and three element letters on sight.

What if a Morse message has no spaces at all?

Then several readings may be correct, because the same marks can split into different letters. Try the splits that make real words, and use the context of the whole message to choose between them.

How do I tell a figure from a letter?

By length. Every figure has five elements and no letter has more than four, so a five-element group is a figure, a few punctuation marks or prosigns, or two letters that ran together.

What does a long row of dits mean in a Morse message?

Eight dits, ........, is the error signal: the sender made a mistake and is about to correct it. Ignore the word just before it.