Skip to content
Menu

Total story points across several Jira spaces

Checked against Atlassian’s documentation on 12 October 2026.

Your teams work in several spaces and you want one total: how many story points are planned, how many done, epic by epic, across all of them. On Reddit it reads “Does anyone have a JQL query that shows total story points by epic, rolled up at the epic level?” This page says what Jira can do across spaces by itself, where each way stops, then what Sumline does.

What Jira does by itself

  • A search finds the work, but adds nothing up. JQL such as project in (WEB, PAY, APP) AND issuetype = Story lists every story in those spaces; it shows no total.
  • Two story-points fields. Company-managed spaces use “Story Points”, team-managed spaces “Story point estimate” (Atlassian’s smart values reference), so a list that spans both kinds of space has two columns to add.
  • A board can span spaces. “Jira allows you to create a company-managed board that displays work items from multiple spaces” from a saved filter (Atlassian). On a Scrum board, the Epic report then shows one epic’s complete, incomplete and unestimated story points (Atlassian), one epic at a time and in one estimation field.
  • Plans, on Premium and Enterprise, rolls estimates up across the spaces in a plan; see Jira roll-ups without Premium.

An export (free, a few minutes)

  1. Search across the spaces, for example project in (WEB, PAY, APP) AND issuetype in (Story, Task, Bug).
  2. Add the columns Parent, Story Points and Story point estimate.
  3. Export to CSV (Atlassian), and in the spreadsheet add the two points columns and total them by parent with a pivot table.

Where it stops:

  • A snapshot, as current as the export.
  • Subtasks are your call. If they carry points too, include them and decide which level counts, or the same work is counted twice.
  • Epics are one level. A total per initiative needs the epics’ parents as well.

An Automation rule can keep a total in a field instead, with a lookup over the spaces; the rule must be multi-space or global to see them, and a lookup returns at most 100 work items (Atlassian). The steps are in Sum story points from subtasks to epics in Jira.

With Sumline

Sumline is our app. Its page under Apps › Sumline is one table across the spaces you choose:

  • One row per epic (or per initiative, on Jira Premium) that is open or was done in the last 90 days, with points done / total, original estimate, logged and remaining time, and % complete by estimate and by issue count.
  • Both story-points fields, found automatically in each space, so team-managed and company-managed spaces add up in one table, and children in other spaces are counted.
  • Each estimate counted once, and unestimated work named.
  • Sort by any column, filter by status, assignee or sprint, open a row to the issues under it, and download the rows as Excel or CSV; the downloads add a Space column.
  • The same rows on a dashboard, with the Sumline gadget.

One limit, stated: a table stops at 2,000 issues, counting its rows and everything under them, and says so rather than showing a partial number. Sumline only reads, so the totals are not a Jira field. More in Getting started.

Sumline under Apps: “12 parents rolled up from 62 issues in APP, PAY, WEB”, sorted by % by estimate: two done epics at 100%, nine in progress, and one not started at 0%. Each epic shows its priority, points done of total, assignee, status, original estimate, time logged, remaining time and both percentages, and three carry “unestimated” warnings. Download Excel and Download CSV sit under the table.
Sumline under Apps across the spaces APP, PAY and WEB: twelve epics rolled up from 62 issues, each with points, time and both percentages.

Sumline is in Marketplace review. Get one email the day it’s live.