Jira Backlog: Sort by Priority, Rank and Missing Items

The product owner opens the backlog on Monday, finds the refund bug that support has been chasing, and sets its priority to Highest. The red double arrow appears. The item stays exactly where it was, ninth of ten, below two items marked Low and Medium. At sprint planning the team takes the top six and the refund bug waits another two weeks.

Nothing was broken. In Jira the backlog has an order of its own, and Priority is not it. We tested how that order works on a Jira Cloud scrum board on 7 November 2026, along with the questions people search for most: sorting by priority, ranking that is disabled, and items that do not show up.

What the Jira backlog is

The 2020 Scrum Guide defines the product backlog as “an emergent, ordered list of what is needed to improve the product”. The word that matters is ordered. The guide makes ordering the Product Owner’s job (“Ordering Product Backlog items”) and sizing the developers’ job (“The Developers who will be doing the work are responsible for the sizing”).

Jira gives that list a screen. Which screen you get depends on the board:

On a kanban board Atlassian is frank about the first-column habit: it works “as long as there are only a few work items in your backlog. As your backlog grows, viewing and scrolling through these work items can become difficult.” How to switch the kanban backlog on is in our Jira kanban board guide. In a team-managed space you add it yourself: select + in the space navigation, then Backlog, then Add to navigation. Turning it off later moves every item back onto the board, “placed in the appropriate column based on their current status”.

Rank decides the order, Priority does not

Every work item on a Jira Software board has a Rank: a text value such as 0|i000e0: that Jira compares to sort the list. The backlog, the board columns and any JQL that ends in ORDER BY Rank all read the same value. Priority is a separate field that the backlog only displays as an icon.

We checked this on our scrum board, PAY, with five new tasks:

• New items from the API or an automation land at the bottom. Our five tasks, created through the REST API with priorities Highest, Low, Medium, Lowest and High, took the last five places in creation order. Items made in the Create dialog behave differently: Atlassian says they go “under the currently selected work item, or at the top of the backlog if no work item is selected”.
• Priority does not move anything. We changed PAY-37 from Lowest to Highest. It kept its place, ninth of the ten items on screen, with the same rank value, below PAY-35 (Low) and PAY-36 (Medium):

• Ranking moves it. One rank call (the API’s version of a drag) put PAY-37 above the first item. Its rank became 0|i000dz:i, a value between its new neighbours, and no other item changed. That is why ranking is cheap: Jira rewrites one value, not the whole list.

In the backlog you rank by dragging, or by right-clicking an item to send it “to the top or the bottom of the Backlog”. The API moves up to 50 items in one call (PUT /rest/agile/1.0/issue/rank), and Atlassian’s API reference ties ranking to the schedule issue permission, so someone who can edit a field may still be unable to reorder.

So which one should drive planning? Use Priority for urgency that other people read: support, an incident queue, a filter for “all Highest bugs”. Use Rank for “what we do next”. When the two disagree, the team works from Rank, because that is what the top of the backlog shows.

Can you sort the Jira backlog by priority?

The backlog follows the board’s saved filter. Change the filter’s ordering and the backlog changes with it. We set the PAY board’s filter to project = "PAY" ORDER BY priority DESC. The Backlog section came back sorted Highest, High, Medium, Low. Then we tried to drag an item:

Nothing moved. Jira’s message: “Work items can’t be manually ranked. The board’s filter query isn’t sorted by rank. To rank work items manually on your board and backlog, sort them by ascending rank.” This is the “jira ranking is disabled” people search for: once the filter is ordered by anything else, manual ranking is off on that board and on its backlog. The fix is the one in the message: end the filter with ORDER BY Rank ASC again, which is what we did.

Three ways to get the priority view without breaking the board:

1. Rank in priority order once. Drag the Highest items to the top and the Lowest to the bottom. You do it once; after that, new urgent items are ranked by whoever triages them.
2. A second view sorted by priority. Run the backlog query below, ordered by priority and then by rank, as a saved filter. You read it by priority; the board stays on Rank.
3. Group instead of sort. A board that shows one column per priority and keeps the rank inside each column gives you both at once. That is what our Priority Board does (more below).

Why work items are missing from your backlog

Atlassian lists four conditions. An item is in the scrum backlog only if “the work item isn’t a subtask, the work item matches the board’s saved filter, the work item’s status maps to one of the board’s columns (but not the Done column), and there’s at least a status being mapped to the right-most column.” The last one surprises people: “If you map all the statuses to the first (To Do) column, you won’t see any work items in the Scrum backlog.”

The checklist we use when someone says “my item disappeared”:

Two of these we saw happen. Moving PAY-38 to Done took it out at once (the API’s count went from 11 to 10). A sub-task we added under PAY-34 showed up in the API’s backlog endpoint, which then listed 11 items, while the Backlog screen said 9 work items: Jira’s REST endpoint /rest/agile/1.0/board/{id}/backlog includes sub-tasks and the screen does not. If a script or an app counts your backlog, it may not agree with the number on the screen.

The kanban backlog has its own rule: an item is visible only if its status “maps to the kanban backlog column and the next column on the kanban board”. Unlike the scrum backlog, it shows a sub-task “slightly indented after its parent task” and lets you move it on its own.

The backlog as one JQL query

There is no “backlog” keyword in JQL, but you can rebuild the scrum backlog. This query returned exactly the ten items of our Backlog section, in the same order:

project = PAY
  AND sprint is EMPTY
  AND statusCategory != Done
  AND issuetype not in (Epic)
  AND issuetype not in subTaskIssueTypes()
ORDER BY Rank ASC

Without the two issuetype lines the result also had three epics and a sub-task. Replace PAY with your project and check two things before you trust it: that your Done column holds only statuses of the Done category, and that the board filter has no extra conditions (a label, a component). If it does, add them here too.

With the query you can save the backlog as a filter, put it on a dashboard, subscribe a stakeholder to it, or export it: in Jira’s work item search (All work), More actions, Export. Change the last line to ORDER BY priority DESC, Rank ASC for the priority view from the previous section. More queries for planning are in our Jira epic vs story vs task and Jira dashboard guides.

Backlog, sprint, board: which is which

The Scrum Guide’s sprint backlog is narrower than Jira’s sprint: it is “composed of the Sprint Goal (why), the set of Product Backlog items selected for the Sprint (what), as well as an actionable plan for delivering the Increment (how)”. Jira’s sprint holds the items and the goal; the plan is up to the team.

Work goes back to the backlog when a sprint ends. Atlassian: “If the sprint has incomplete work items”, you choose Backlog, any future sprint or a new sprint, and a work item counts as complete only “once they have been moved to the last column on the board”. All sub-tasks must be Done before the sprint can be completed.

One more rule from Atlassian: while a sprint is active “you can still plan subsequent sprints in the backlog, but you won’t be able to start them until the active sprint is completed”, unless parallel sprints are switched on. And starting a sprint needs space admin rights “for all spaces that match the board’s filter”.

Backlog refinement in Jira, step by step

The Scrum Guide describes refinement as “the act of breaking down and further defining Product Backlog items into smaller more precise items”, adding “a description, order, and size”. In Jira that maps onto four things you can do without leaving the backlog:

1. Order. Drag, or right-click and send to the top or the bottom. Items “deemed ready for selection” should sit at the top before sprint planning.
2. Split. Right-click, Split work item. Atlassian lists what does not travel: “work log, comments, work item history, and links”, and the new item’s status returns “to the first step of the corresponding workflow”. The original is linked to the new one.
3. Size. Select an item and press E to edit; J and K move through the list. Story Points are available only on stories and epics by default.
4. Edit in place. Summary, assignee, estimate, parent and priority can be changed on the row. Bulk changes go through the checkboxes and Edit.

A quick filter per epic, label or assignee keeps a long backlog readable. If refinement turns into a debate about what is an epic and what is a story, our epic vs story vs task guide has the rules we tested.

Apps for the backlog

Jira’s own backlog does ordering well and stops there. To see what teams add on top, we searched the Atlassian Marketplace API on 7 November 2026 for Jira Cloud apps about the backlog (ten searches, from “backlog” and “prioritization” to “wsjf” and “story map”). Of 159 results, 71 have the backlog as their main job; the rest are portfolio, roadmap, portal or security tools that only mention it.

Three things stand out. Story maps and estimation take 79% of the installs: teams buy help with what to build and how big it is, not with the list itself. Scoring apps are many and small; 16 of the 20 have fewer than 100 installs. And ordering tools barely register (227 installs across nine apps, ours included), which fits what we found above: Rank already orders the backlog, and what people search for is how to see it by priority without breaking that order. 57 of the 71 apps are free, or free for up to 10 users.

If you only want scores, try them before you install anything: our RICE & WSJF calculator runs both methods on the same list in your browser.

Where our app fits

Priority Board answers the sorting problem from this article without touching your board filter. It shows a project as one column per priority, plus a No priority column while some items have none. Dragging a card to another column sets the real Priority field, as you; dragging it up or down changes its Rank, so the order you set inside a column is the order your Jira backlog and boards show. A Sprints view lists the backlog and the active and future sprints in Jira’s order and lets you drag between them.

It does not score items: there is no RICE or WSJF inside. For that, our free RICE & WSJF calculator ranks a list both ways in your browser and suggests a Jira priority for each item. Priority Board is free for up to 10 users; install it from the Atlassian Marketplace.

How we tested

• Our Jira Cloud site, 7 November 2026, board 4 “PAY board”, a company-managed scrum board with filter project = "PAY" ORDER BY Rank ASC, five columns (To Do, In Progress, In Review, QA, Done) and one active and two future sprints.
• Work items PAY-34 to PAY-38 created with POST /rest/api/3/issue, sub-task PAY-39 under PAY-34. Order read from GET /rest/agile/1.0/board/4/backlog and from the Backlog screen after a reload.
• Priority changed with PUT /rest/api/3/issue/PAY-37; rank changed with PUT /rest/agile/1.0/issue/rank; PAY-38 moved to Done with its workflow transition.
• Board filter changed to ORDER BY priority DESC, a drag tried in the browser, then the filter restored. Not tested: mapping every status to the first column and the Create dialog placement; both are quoted from Atlassian.

Sources

• Our Jira Cloud site, 7 November 2026: board PAY, work items PAY-34 to PAY-39.
• Atlassian Support: Use your scrum backlog, Use your kanban backlog, Enable the backlog (team-managed spaces).
• Atlassian Support: Complete a sprint (company-managed spaces).
• Atlassian Marketplace REST API v2 (/rest/2/addons, distribution.totalInstalls, Cloud price lists), read 7 November 2026; our grouping by main job.
• Atlassian developer documentation: Jira Software Cloud REST API, Issue (rank and backlog endpoints).
• The 2020 Scrum Guide.