Jira Markdown: Formatting Text in Jira Issues

Which Jira Markdown works depends on the edition you're using. Jira Cloud converts shortcuts such as **bold** and # Heading into rich text as you type, but stores the result as its own document format. Jira Data Center uses wiki markup with different symbols entirely. This guide gives you a three-column cheat sheet, a straight answer on pasting Markdown, and a workflow for drafting issues in real Markdown first.

Does Jira Use Markdown?

Not in the way GitHub does. There are three products people mean when they say "Jira" and each treats formatting differently.

Jira Cloud has a rich-text editor. Atlassian's Markdown and keyboard shortcuts page lists the Markdown shortcuts it recognises: when you type **bold** the editor replaces the asterisks with bold text. The issue is stored in Atlassian Document Format (ADF), a JSON structure, not as Markdown. Atlassian states it plainly: you can copy raw markdown into the editor, but you can't convert content back into raw markdown.

Jira Data Center (and the discontinued Jira Server) uses Jira wiki markup, also called text formatting notation. Bold is *text*, headings are h1., and code goes inside {code} tags. Markdown syntax does nothing here.

Bitbucket is the only Atlassian product in the search results that renders real Markdown, in READMEs and pull request descriptions. The Bitbucket guide that ranks for Jira queries is about the wrong product.

So "jira markdown" is shorthand for two different things: shortcuts that trigger rich text in Cloud, and a separate wiki syntax in Data Center. The cheat sheet below covers both.

Markdown to Jira Cheat Sheet

The left column is standard Markdown. The middle column shows what the Jira Cloud editor does when you type those characters, from Atlassian's shortcut list. The right column is the Data Center wiki markup from the text formatting notation help that ships with every Data Center instance.

Markdown	Jira Cloud editor (type it)	Jira Data Center wiki markup
**bold** or __bold__	Converts to bold	*bold*
*italic* or _italic_	Converts to italic	_italic_
~~struck~~	Converts to strikethrough	-struck-
`code`	Converts to monospace	{{code}}
# Heading to ###### Heading	Converts to heading 1 to 6	h1. Heading to h6. Heading
- item or * item	Starts a bullet list	* item (** to nest)
1. item or 2) item	Starts a numbered list	# item (## to nest)
[] task	Creates an action item	No equivalent
> quote	Converts to a block quote	bq. quote or {quote}...{quote}
--- or ***	Inserts a horizontal rule	----
```	Starts a code block	{code:java}...{code}
[text](url)	Converts to a link	[text|url]
	Inserts an image	!image.png!
| a | b | pipe table	Not in the documented shortcuts	||Header|| then |cell|
Preformatted text	Code block	{noformat}...{noformat}

Three rows deserve attention. Bold is the classic trap: a single asterisk is bold in Data Center and italic in Markdown, so *important* means different things in the two systems. Links flip the separator from parentheses to a pipe. And Jira Cloud's shortcut list has no table syntax; the Atlassian Community reports typing pipes creates a table in some editor versions, but we couldn't confirm it in the documentation, so use the /table command instead.

How Do You Format Text in Jira Cloud With Markdown Shortcuts?

Type the Markdown characters and keep typing. The conversion happens when the closing character or the space after it is entered, so **done** becomes done the moment you type the second pair of asterisks. Headings convert when you type # at the start of a line, and 1. starts a numbered list.

Two behaviours differ from a Markdown editor. Typing ** at the start of a line doesn't create a nested list; Atlassian's docs say so explicitly, and you nest with Tab instead. And once a shortcut has converted, the source characters are gone. Pressing Ctrl+Z straight after a conversion usually restores the literal characters, which is the quickest way to type a real asterisk.

Two shortcuts have no Markdown parent but are worth knowing. Typing [] followed by a space at the start of a line creates an action item, Jira's checkbox, which is the closest thing to a GFM task list. Typing : followed by an emoji name, such as :warning:, inserts the emoji. Anything else, from tables to mentions to status lozenges, comes from the / quick-insert menu, which is also the fallback when you can't remember whether a shortcut exists.

Code blocks are the top formatting need in issues, and the shortcut is the same three backticks you'd use anywhere. Type ``` on a new line, and the editor inserts a code block with a language picker. The Markdown code block guide covers the fenced syntax in general; in Jira Cloud the language is picked from a dropdown rather than typed after the fence.

How Do I Paste Markdown Into Jira?

This is the question the People Also Ask box asks and nobody answers directly. Here is what Atlassian's documentation and the Atlassian Community threads say, as of September 2026.

Pasting into Jira Cloud. Atlassian's shortcut page says you can copy raw Markdown into the editor. Community threads report that simple marks such as headings, bold, and list markers usually convert on paste, while pipe tables and nested lists arrive as literal text.

Results vary by editor version, so treat a paste as a starting point. For a short comment, paste and fix by hand. For a long description, rebuild tables with /table, and put anything that must stay exact inside a code block.

Pasting into Jira Data Center. Markdown stays literal. **bold** shows the asterisks, # Heading shows the hash, and a pipe table shows as text. You have to translate to wiki markup using the cheat sheet above before pasting, or use the visual editor toolbar after pasting plain text.

Jira to Markdown: Getting Text Back Out

There's no export to Markdown from either edition, which surprises people who assumed the Markdown Jira Cloud understands was being stored. Jira Cloud's REST API returns descriptions as ADF JSON, documented in the Atlassian Document Format structure reference. Converting ADF to Markdown means a script you write yourself or a marketplace app. Older Data Center REST endpoints return the wiki markup string, which is at least readable and can be translated with the cheat sheet in reverse.

For a one-off, the rendered issue is the easiest source. Copy the description from the browser and paste it into the HTML to Markdown converter, which turns the formatted text into headings, lists, and code blocks. If your browser passes only plain text on paste, copy the description's HTML from the developer tools instead. Either way, check the result: Jira's action items and mentions have no Markdown equivalent and come through as plain text.

Draft the Issue in Markdown First

We prefer drafting long issue descriptions in a real Markdown editor and pasting the result. The draft survives even if Jira eats the formatting. The editor below has a bug report template. Write it here, then copy the text into Jira Cloud and check which marks converted, or translate the marks for Data Center.

The log excerpt is a block quote rather than a fenced block on purpose. A quote pastes as readable text you can select and turn into a code block with the toolbar, which is more predictable than hoping a pasted fence converts. For Data Center, wrap that excerpt in {noformat} before pasting.

Jira Data Center Wiki Markup in Practice

If you're on Data Center, the same bug report looks like this:

h1. Login fails after password reset

h2. Steps to reproduce
# Request a password reset from the sign-in page
# Open the emailed link and set a new password
# Sign in with the new password

h2. Expected
User lands on the dashboard.

h2. Actual
*Error 401* is returned and the user is sent back to sign-in.

h2. Environment
* Build: 4.12.0 (commit a1b2c3d)
* Browser: Edge 128 on Windows 11

{noformat}
401 Unauthorized: token expired at 2026-09-11T08:41:02Z
{noformat}

Tables use double pipes for the header row and single pipes for cells, all on one line per row:

||Field||Value||
|Build|4.12.0|
|Browser|Edge 128|

One thing to check before blaming the syntax: the field's renderer. Atlassian's support article on advanced formatting not working in Jira text fields explains that {code} and {noformat} only render when the field uses the Wiki Style Renderer. A field set to the default text renderer shows the braces literally.

Common Jira Markdown Mistakes

Using single asterisks for emphasis on Data Center. *text* is bold in wiki markup, not italic. Use underscores for italic and you'll get what you expected on both editions.

Pasting a Markdown table into Jira Cloud. The pipes stay as pipes. Insert a table with /table and paste cell contents, or convert the rows to ||header|| format for Data Center. The Markdown table guide shows the pipe syntax if you need to go the other way for a README.

Expecting the description to round-trip. Text that went into Jira Cloud as Markdown comes back out as ADF JSON. Keep the Markdown source in a repo or the editor if you'll need it again.

The acknowledged limitation of all this: Atlassian can change which shortcuts fire and how paste behaves without notice. The editor has changed several times since the new issue view arrived. Test with a throwaway issue before relying on a paste workflow for a 40-line description. The same "looks like Markdown, isn't Markdown" pattern shows up in the Slack Markdown table post. The Discord Markdown cheat sheet covers a platform that gets closer to the real thing.

Markdown in Jira FAQ

Knowing which Jira Markdown works saves you from typing a heading and getting a hash. Jira Cloud converts the common shortcuts as you type, Data Center wants wiki markup, and neither stores Markdown. Draft the description in the editor, then paste into Cloud or translate for Data Center. Keep the source file so the next update starts from text instead of from ADF JSON.