Checked against Atlassian docs on 17 Oct 2026

How to export more than 10,000 Jira issues

Short answer: Jira Cloud's CSV export stops at 10,000 issues. To get more, split your JQL by created date into batches under 10,000, export each batch, and merge the files. It is free and takes some care. Or use an app that splits the files for you.

Which option fits you

Your situation	Best fit
One export, a bit over 10,000, no budget	Split by created date and merge (free, steps below)
You live in Google Sheets	Jira Cloud for Sheets with two `=JIRA()` calls
You have a developer and need every field	A REST API script
You need templates or Word documents	A template app from the Marketplace
You repeat a large export every week or month	A scheduling app, such as Haulcopy

Where the 10,000 limit came from

On 19 March 2025, Atlassian raised the CSV export from Jira's issue search from 1,000 to 10,000 issues. The export runs in the background, and Jira sends an in-product notification with a download link when it is ready.

The change did not touch the REST API. Old scripts that use /rest/api/3/search stopped working when Atlassian removed that endpoint (see the API option below). The pager/start URL trick was deprecated on Cloud in March 2026.

Still seeing 1,000? You may be on Data Center: why your Jira export stops at 1,000.

Split your JQL and merge the files (free)

Adapted from Atlassian's own method. No app needed.

1. Count the full search

Go to Filters, then View all work items. Run your JQL with today's date as an upper bound, and note the total:

project = MYPROJECT AND created < "2026-10-17" ORDER BY created DESC

The fixed end date freezes the set, so new issues can't change the count while you work.

2. Split on the created date

Field	Use it?
created	Yes. It never changes after an issue is created
updated	Avoid. It changes while you export
status	Avoid. Issues move while you export

Atlassian calls created "the most predictable field". Avoiding updated and status is our advice. Atlassian lists them as options.

3. Write batches that don't overlap

Start each range with >= and end it with <, so no issue fits two batches. Keep the same upper bound in the newest batch:

Batch 1: project = MYPROJECT AND created >= "2026-01-01" AND created < "2026-10-17" ORDER BY created DESC
Batch 2: project = MYPROJECT AND created >= "2025-06-01" AND created < "2026-01-01" ORDER BY created DESC
Batch 3: project = MYPROJECT AND created < "2025-06-01" ORDER BY created DESC

Busy months need shorter ranges.

4. Check the counts

Run each batch. Every batch must show fewer than 10,000 results. If one is too big, halve its range. Then add the batch counts. The sum must match step 1: lower means a gap, higher means an overlap.

5. Export each batch

Open More actions (three dots), choose Export, and pick the same CSV option for every batch. Atlassian warns that all fields exports are slow and produce large files, so set your columns and use current fields or filter fields.

6. Merge and check

Put only these CSV files in one folder. In Excel, choose Data, Get Data, From File, From Folder, then Combine & Load. In the Combine dialog, set File Origin to UTF-8 and check Data Type Detection, so keys and dates stay as they are.

Atlassian's warning: "we recommend that you don't upload the file to apps like Google Sheets or Microsoft Excel as the file may execute macros." Importing through Power Query, as above, is safer than opening the file directly.

Finally, run Remove Duplicates on the Key column. The row count should match step 1.

Other options

Jira Cloud for Sheets

Atlassian's Sheets add-on handles up to 10,000 issues per import. For more, call =JIRA() twice on a sorted query, with offsets 0 and 10000. You still combine results, but in one sheet.

The REST API

Atlassian's 10,000 limit is for the export in Jira's screens. The API pages through results instead, with /rest/api/3/search/jql and a nextPageToken. You build and maintain the script.

Details for developers

The old /rest/api/3/search endpoint with startAt has been removed (Atlassian KB, Dec 2025).
The JQL must be bounded: a real filter, not only an ORDER BY.
It returns only issue ids unless you ask for fields.
It returns no total. Use /rest/api/3/search/approximate-count.
The last page has no nextPageToken.

Marketplace apps

Several apps export Jira issues to files, for example Better Excel Exporter (Midori), Xporter (Xblend, formerly Xpand IT) and Advanced Export. Limits, formats and prices differ, so check each listing against your export size.

How Haulcopy handles searches over 10,000

Haulcopy is a Jira Cloud app that exports a JQL search to Excel, CSV or PDF.

In our test, 10,000 issues with 50 fields exported to a formatted Excel file in about 2 minutes.
Above 10,000 issues, files split automatically into files of up to 10,000.
No template language and no watermark. Up to 1,000 schedules.
Your data stays in Atlassian. Haulcopy has no servers of its own.
Built by a solo developer in Portugal.

Coming soon to the Atlassian Marketplace

FAQ

What is the Jira Cloud export limit?

10,000 issues per CSV export from the issue search, since March 2025. For Word, Atlassian recommends about 150 at a time.

Does the pager/start URL trick still work?

Not on Cloud. Atlassian deprecated it in March 2026. Use JQL batches.

Does Atlassian's Excel export give me an .xlsx file?

No. "Export Excel CSV" produces a CSV. The exception is "Export XLSX (form data)" for Jira Service Management forms. A formatted .xlsx is a different path: export Jira issues to Excel.

Can the REST API return more than 10,000 issues?

The 10,000 limit is for the export in Jira's screens. The API pages with nextPageToken, but you write and run the script.

Does Haulcopy put more than 10,000 issues in one file?

No. Above 10,000 issues, files split automatically into files of up to 10,000.

Sources: Atlassian docs and Microsoft Support, checked 17 Oct 2026.