# Export Jira Data and Confluence Spaces: Complete Guide

> Export Jira data and Confluence spaces from Cloud: site backups vs CSV, Word, HTML and XML, the 10,000-item CSV limit, changelogs and 60-day retention.

Canonical: https://origindatapartners.com/guides/export-jira-confluence-data
By: Origin Data Partners
Status: Published
Published: 2026-11-10
Updated: 2026-11-14
Sources checked: 2026-11-10

## Quick answer

To export Jira data completely, use Settings, System, Backup manager and create a backup for cloud with attachments; CSV exports from search are capped at 10,000 work items. To export a Confluence space, open Space settings, General, Export space, and choose XML or CSV for everything. A full Confluence site backup lives in Confluence administration under Data management.

## Key points

- The Jira site backup is the only complete copy: work items, comments, boards, sprints, users and configuration, with attachments if ticked.
- Search exports are for slices: CSV up to 10,000 work items, Word and XML with descriptions and comments.
- Change history needs the REST API changelog endpoints if you want it outside a backup.
- Cancelling gives 15 days of access, then 60 days of retention; unsubscribing from every Jira app deletes Jira data at once.

## How to export Jira data and Confluence spaces: which export to use

Atlassian Cloud has two kinds of export. A site backup copies everything in one app (Jira or Confluence) for restore or migration. A content export copies a slice, such as a Jira search or a Confluence space, into CSV, Word, HTML, XML or PDF. For an archive you usually want the backup first and targeted exports second. Organization admins can also block exports in particular app instances or with certain classification levels ([Atlassian](https://support.atlassian.com/jira-cloud-administration/docs/export-issues/)), so check that first if an option is missing.

This guide is for engineering managers, Jira and Confluence admins and owners who want a complete copy of their issue tracker and wiki. If you want to know whether the history is worth licensing, [get a free estimate](https://origindatapartners.com/check-fit) from company-level answers; nothing needs exporting first.

| Export method | What it includes | Limits | Who can run it |
| --- | --- | --- | --- |
| Jira backup for cloud (Settings > System > Backup manager) | Work items and field content, comments, boards and sprints, users and groups, space configuration, attachments if ticked | No automation flows, third-party app data or Assets; 48 hours between backups with attachments; recommended within 25 GB of data | Jira admins |
| Jira search export | CSV (all or current fields), Word, Excel, XML or Google Sheets for the work items a search returns | CSV up to 10,000 work items; Word best around 150 work items at a time | Anyone who can see the work items |
| Jira REST API changelogs | Dated history of field changes for each work item | Up to 1000 issues per bulk changelog request | Users with browse permission |
| Confluence space export (Space settings > General > Export space) | PDF, CSV, HTML or XML of a space; CSV includes attachments and comments by default | HTML leaves out page comments and blog posts; PDF leaves out comments | Space admins |
| Confluence site backup (Data management > Backup manager) | Pages, whiteboards, attachments if selected, selected user and group settings, team calendars | One backup file at a time, link valid 14 days; no databases or trashed spaces; up to 30 GB plus 800 GB of attachments | Confluence admins |

Source: [Atlassian Support: Export data from Jira Cloud](https://support.atlassian.com/jira-cloud-administration/docs/export-issues/)

## Export Jira data with a site backup

A Jira Cloud backup covers "Jira work items and field content (including system and custom fields)", board and sprint data, users and user group settings, comments, media files if selected, and space configuration ([Atlassian](https://support.atlassian.com/jira-cloud-administration/docs/export-issues/)). It is the most complete Jira export there is.

- Select Settings, then System. In the Import and Export section, select Backup manager.
- Under Backup for cloud, select Create backup for cloud, and tick the option to include work item attachments, user avatars and space logos.
- When the backup completes, select Download backup file. Keep it with a note of the date.
- Allow 48 hours between backups that include attachments. Backups without them can run at any time.
- Export automation flows separately; the backup leaves out automation, third-party app data, Opsgenie features and Assets for Jira Service Management.
- Unzip and check: database data is stored in activeobjects.xml and entities.xml, with attachments, avatars and logos in their own folders.
- Keep backups fresh if you may restore them. From 22 January 2026 Backup Manager only restores backups that are 30 days old or less.

Source: [Atlassian Support: Export data from Jira Cloud](https://support.atlassian.com/jira-cloud-administration/docs/export-issues/)

## Export Jira work items to CSV, Word or XML, and keep their history

For a readable slice, run a search and use the More actions menu (three dots), then Export. The options are CSV, Google Sheets, Microsoft Word, Microsoft Excel and XML; CSV comes as all fields or current fields ([Atlassian](https://support.atlassian.com/jira-software-cloud/docs/export-search-results/)). The Word export includes the description, comments and all other work item data, and Atlassian recommends exporting around 150 work items at a time.

CSV has a ceiling. Exporting up to 10,000 work items with the asynchronous Export CSV feature is supported, and exporting all fields "can significantly slow down the process and result in large files that are difficult to manage" ([Atlassian](https://support.atlassian.com/jira/kb/export-over-10-000-work-items-in-jira-cloud/)). Split larger exports by project or date with JQL.

Neither export gives the full change history in a tidy form. The REST API's bulk changelog endpoint "Returns a paginated list of all changelogs for given issues sorted by changelog date and issue IDs", for up to 1000 issues per request ([Atlassian Developer](https://developer.atlassian.com/cloud/jira/platform/rest/v3/api-group-issues/)). That is where status changes, reassignments and estimate changes live, with their dates.

Source: [Atlassian Support: Export search results (Jira Cloud)](https://support.atlassian.com/jira-software-cloud/docs/export-search-results/)

## How to export a Confluence space

Confluence content "can be exported to various formats, including Microsoft Word, HTML, CSV, PDF, and XML", and space admins can export an entire space or selected content ([Atlassian](https://support.atlassian.com/confluence-cloud/docs/export-content-to-word-pdf-html-and-xml/)).

- Next to the space name in the sidebar, select More actions, then Space settings. Open the General menu and select Export space.
- Choose XML or CSV for an archive. CSV exports everything you can view, including attachments and comments by default; XML is the format for importing into Confluence Data Center.
- Run the export as a site admin. Only content visible to you is exported, except when a site admin exports via CSV or XML, in which case all content is exported.
- Avoid PDF and HTML for an archive: comments are never included in PDF exports, page comments are not exported in HTML, and blog posts are left out of both.
- Note that only the published version of each page is exported, and team calendars are not included.

Source: [Atlassian Support: Export a space's content as Word, PDF, or data files](https://support.atlassian.com/confluence-cloud/docs/export-content-to-word-pdf-html-and-xml/)

## Back up a whole Confluence site

For every space at once, open Confluence administration, then Settings, Data management, Backup manager, select Backup attachments and choose Create backup for cloud. "Only one backup file is stored at a time. Any existing backup is overwritten by the new one." The download link is available for 14 days ([Atlassian](https://support.atlassian.com/confluence-cloud/docs/create-a-site-backup/)).

The backup includes pages, whiteboards, selected user and group settings, team calendar data and attachments if selected, including those in the archive or trash. It leaves out soft-deleted spaces until you restore them from the trash, and databases keep their place in the tree but not their content. A backup file can be up to 30 GB in total, plus 800 GB of attachments ([Atlassian](https://support.atlassian.com/confluence-cloud/docs/create-a-site-backup/)).

Source: [Atlassian Support: Export a data backup file of your Confluence Cloud instance](https://support.atlassian.com/confluence-cloud/docs/create-a-site-backup/)

## Retention traps: cancelling Jira and Confluence

After a paid subscription is cancelled and its period ends, "you'll be able to access your site for 15 more days." The site is then deactivated and the data kept for 60 days on Free, Standard, Premium and Enterprise plans (15 days for trials) before it is permanently deleted. One rule removes even that cushion: "If you unsubscribe from all Jira apps on your site, your Jira data will be deleted immediately." ([Atlassian](https://support.atlassian.com/security-and-access-policies/docs/track-storage-and-move-data-across-products/))

Export Jira and Confluence before touching subscriptions, and keep a site admin account until the files are checked. The [checklist for exporting before closing](https://origindatapartners.com/guides/export-company-records-before-closing) puts these windows next to Google Workspace, Slack, Microsoft 365, Salesforce and Zendesk.

Source: [Atlassian Support: data retention after cancellation](https://support.atlassian.com/security-and-access-policies/docs/track-storage-and-move-data-across-products/)

## Keep the export complete and linked to code

A key link in an issue tracker points outside it. Atlassian's own guidance is to "Include the work item key in the commit message to link the commit to your Jira work item" ([Atlassian](https://support.atlassian.com/jira-software-cloud/docs/reference-issues-in-your-development-work/)), and to use the key in branch names and pull request titles. If your team did this, the Git history of each repository holds the other half of every ticket.

- Export the repositories as full mirror clones alongside the Jira backup; the work item keys in commit messages join the two.
- Keep work item keys, IDs, parent and sub-task links, issue links, sprint and version fields, created, updated and resolved dates.
- Keep Confluence page IDs and the links from pages to work items; design documents and post-incident reviews often point at the tickets they explain.
- Record the date, the site and the options used for each backup and export in the [data inventory template](https://origindatapartners.com/blog/company-data-inventory-template).

Source: [Atlassian Support: Reference work items in your development work](https://support.atlassian.com/jira-software-cloud/docs/reference-issues-in-your-development-work/)

## Why issue trackers linked to code matter to AI developers

AI developers building coding and engineering agents need to see how real software work is planned, argued over, built, reviewed and shipped. A Jira history with its comments and changelog, joined to the commits and pull requests that closed each item and the Confluence pages that explained the design, shows the whole loop with the outcome. In Origin's experience as an independent data partner, that linkage is what separates a valuable engineering archive from a pile of tickets.

In Jira the outcome signal is the resolution and resolved date, the status changelog that shows how long each step took and how often work was reopened, and the commits and pull requests that closed each item. Confluence adds the design pages and post-incident reviews that explain why. Code and workspace archives from small companies do sell: [MIT Technology Review](https://www.technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models/) reported a founder netting roughly $25,000 for a code base, with around $15,000 more expected for workspace data.

See [licensing source code](https://origindatapartners.com/guides/sell-source-code-ai-training), [licensing software documentation](https://origindatapartners.com/blog/license-software-documentation), [incident reviews](https://origindatapartners.com/blog/license-software-incident-reviews), [workflow data for AI agents](https://origindatapartners.com/guides/license-workflow-data-for-ai-agents) and [what company data is worth](https://origindatapartners.com/guides/company-data-value).

Source: [MIT Technology Review: AI labs are buying data from dead startups (Oct 2026)](https://www.technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models/)

## Personal data, customer contracts and who owns Jira and Confluence data

Atlassian's customer agreement says that as between the parties, "Customer owns all intellectual property and other rights in Customer Data and Customer Materials provided to Atlassian or used with the Products", and that after the agreement ends Atlassian will delete Customer Data in accordance with its documentation ([Atlassian Customer Agreement](https://www.atlassian.com/legal/atlassian-customer-agreement)). Ownership of what is inside is a separate question: client projects, contractor work and customer bug reports may belong to, or be confidential to, someone else.

Personal data in Jira and Confluence sits in assignee, reporter and watcher fields, in @mentions in comments, in customer bug reports pasted with names and emails, and in user profiles and avatars inside a site backup. HR, people and security spaces in Confluence should stay out entirely; suppliers of workplace archives already "exclude HR records", [MIT Technology Review](https://www.technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models/) reported.

- Check client and contractor agreements before including projects built for others, and leave out work under someone else's confidentiality terms.
- Remove credentials, API keys and customer data pasted into tickets and pages.
- Replace names and emails of staff and customers with stable tokens, so work items still link to the same person without naming them.
- See the [de-identification guide](https://origindatapartners.com/guides/de-identify-company-data-before-licensing) and [what company data can be licensed](https://origindatapartners.com/guides/what-company-data-can-you-sell-to-ai).

Source: [Atlassian Customer Agreement](https://www.atlassian.com/legal/atlassian-customer-agreement)

## Next step

Run a Jira backup for cloud with attachments and a Confluence site backup today, download both, and export changelogs through the API if the history matters. Keep a mirror clone of every repository next to them. Then note the years covered, the number of work items and pages, and the products they concern, and [get an estimate](https://origindatapartners.com/check-fit) from that description. Support teams on Zendesk can follow the [Zendesk ticket export guide](https://origindatapartners.com/guides/export-zendesk-tickets).

This page maps the vendor documentation and public rules as of November 2026. Plans and menus change, so confirm each step in your own admin console, and take contract and privacy questions to your lawyer.

Not sure whether a buyer or an offer is sound? Origin checks who the buyer is, what the license allows and how the price compares with published deals before anything is signed. [Get a free estimate and a second opinion](https://origindatapartners.com/check-fit).


## Questions and answers

### How do I export all data from Jira Cloud?

Use Settings, System, Backup manager, then Create backup for cloud with the attachments option ticked, and download the file when it completes. It covers work items, fields, comments, boards, sprints, users and configuration, but not automation flows or third-party app data.

### How many issues can I export from Jira to CSV?

Jira Cloud supports exporting up to 10,000 work items with the asynchronous Export CSV feature in the issue navigator. Split larger sets with JQL by project or date, or use the site backup or the REST API instead.

### How do I export a Confluence space?

In the space sidebar, select More actions, Space settings, then General, Export space, and choose PDF, CSV, HTML or XML. For an archive, use XML or CSV as a site admin, because CSV includes attachments and comments and site admins export all content.

### Does a Jira export include comments and attachments?

The Jira backup for cloud includes comments, and attachments if you tick that option. Search exports to Word and XML include descriptions and comments; CSV exports are field-based and do not carry the attachment files.

### How do I export Jira issue history?

Use the Jira Cloud REST API changelog endpoints, which return each work item's changes sorted by date, for up to 1000 issues per bulk request. The site backup also preserves the data for a restore, but the API gives the history in a usable form.

### How long does Atlassian keep data after I cancel?

You keep access for 15 days after the subscription period ends, then the site is deactivated and data is kept for 60 days on Free, Standard, Premium and Enterprise plans, or 15 days for trials, before permanent deletion. Unsubscribing from every Jira app deletes Jira data immediately.

### What does a Confluence site backup leave out?

Spaces in the trash, the content of databases, and the full set of user and group settings. Only one backup file is kept at a time, and its download link lasts 14 days.

### Can Jira and Confluence data be licensed for AI training?

Often, when the company owns the work and removes personal data, secrets and client-confidential projects first. Issue histories linked to commits and design pages are what AI developers building engineering agents look for.

### How do I migrate Jira or Confluence data to another Atlassian site?

Atlassian's data transfer moves projects, spaces, users and related data from one instance of an Atlassian app to another ([Atlassian](https://support.atlassian.com/organization-administration/docs/copy-product-data/)). From Jira Server or Data Center, the free Jira Cloud Migration Assistant moves the projects, users, attachments and Marketplace apps you select, and Atlassian says it will not overwrite existing data in the cloud site ([Atlassian](https://support.atlassian.com/migration/docs/cloud-migration-methods-for-jira/)). Take a site backup first either way, so you hold a copy outside both sites.

## Sources

- [Atlassian Support: Export data from Jira Cloud](https://support.atlassian.com/jira-cloud-administration/docs/export-issues/)
- [Atlassian Support: Export search results (Jira Cloud)](https://support.atlassian.com/jira-software-cloud/docs/export-search-results/)
- [Atlassian Support: Export over 10,000 work items in Jira Cloud](https://support.atlassian.com/jira/kb/export-over-10-000-work-items-in-jira-cloud/)
- [Atlassian Developer: Jira Cloud REST API, issues (bulk fetch changelogs)](https://developer.atlassian.com/cloud/jira/platform/rest/v3/api-group-issues/)
- [Atlassian Support: Reference work items in your development work](https://support.atlassian.com/jira-software-cloud/docs/reference-issues-in-your-development-work/)
- [Atlassian Support: Export a space's content as Word, PDF, or data files](https://support.atlassian.com/confluence-cloud/docs/export-content-to-word-pdf-html-and-xml/)
- [Atlassian Support: Export a data backup file of your Confluence Cloud instance](https://support.atlassian.com/confluence-cloud/docs/create-a-site-backup/)
- [Atlassian Support: data retention after cancellation](https://support.atlassian.com/security-and-access-policies/docs/track-storage-and-move-data-across-products/)
- [Atlassian Customer Agreement](https://www.atlassian.com/legal/atlassian-customer-agreement)
- [California Civil Code 1798.140 (CCPA definitions)](https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=CIV&sectionNum=1798.140)
- [MIT Technology Review: AI labs are buying data from dead startups (Oct 2026)](https://www.technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models/)
- [Atlassian Support: Transfer data from one Atlassian app instance to another](https://support.atlassian.com/organization-administration/docs/copy-product-data/)
- [Atlassian Support: Cloud migration methods for Jira](https://support.atlassian.com/migration/docs/cloud-migration-methods-for-jira/)

## Related reading

- [Cancel Google Workspace or delete Slack? Export data first](https://origindatapartners.com/guides/export-company-records-before-closing)
- [Can a company sell its source code to AI companies for training?](https://origindatapartners.com/guides/sell-source-code-ai-training)
- [/blog/license-software-documentation](https://origindatapartners.com/blog/license-software-documentation)
- [License workflow data for AI agents: a company owner’s guide](https://origindatapartners.com/guides/license-workflow-data-for-ai-agents)
- [How to Export Zendesk Tickets with Comments and History](https://origindatapartners.com/guides/export-zendesk-tickets)
- [Export Agent Traces: LangSmith, Langfuse, Phoenix and More](https://origindatapartners.com/guides/export-agent-traces)