How to Fix the n8n IMAP Trigger That Stopped Working After an Update
Quick answer
Re-save the IMAP credential, delete and recreate the Email Trigger (IMAP) node, switch Format to Resolved, and set Force Reconnect Every Minutes to 5. n8n 2.40 rewrote the trigger, so the old setup silently stops firing.
What broke, and why
Your Email Trigger (IMAP) workflow ran fine for months. Then you upgraded to n8n 2.4x, and new mail stopped starting executions. Manual runs sometimes return mail, sometimes hang, and the workflow shows no error at all. Nothing in your config changed: the breakage is in n8n, not in your workflow.
n8n 2.40 rewrote the IMAP trigger. The new version no longer reads unread mail that is already sitting in the mailbox when it connects. It only reacts when the server announces a new message, plus it rescans for unread mail after every reconnect. If your mail server never sends that new-mail announcement, the trigger waits forever in silence. That is why the failure looks so strange: no error, no failed execution, mail piling up unread. The official Email Trigger (IMAP) documentation confirms the node listens with IMAP IDLE and polls only when IDLE is unavailable.
The rewrite exposed two more faults. In Simple format, a message whose fetched header is missing gets skipped, and a malformed header can crash the trigger and deactivate the workflow in the background while mail keeps being marked as read. The n8n team confirmed this and shipped the repair in pull request #30375, writing that they had “identified and fixed several bugs in the IMAP Email Trigger node”. A separate provider-specific bug hit STRATO mailboxes, and its fix was merged on October 15, 2026.
The September 24, 2026 community thread about this regression gathered 14 replies and 253 views by mid-October, with users on n8n Cloud 2.40.6 reporting the same silent failure in manual and automatic executions alike.
The fix
Work through these in order; most people only need the first four.
- Re-save the credential. Open the IMAP credential, re-enter the password, and save. Credential handling changed across upgrades, and triggers fail silently on stale credentials.
- Reset the trigger node. Delete only the Email Trigger (IMAP) node and add a fresh one. This clears the last-seen UID watermark, which can reset during an upgrade and make the trigger “succeed” with zero new mail.
- Switch Format to Resolved. In the node settings, change Format from Simple to Resolved. Resolved fetches and parses the full message instead of skipping mail with a missing header.
- Set Force Reconnect Every Minutes to 5. Find it under Options. It defaults to 60 when you enable it, which is why switching it on alone looks like it did nothing. Each reconnect rescans for unread mail the trigger has not handled yet, catching mail even when the server never announces it. Keep Fetch Only New Emails on, or every rescan returns all your unread mail again.
- Update n8n. If your mailbox is on STRATO, or you are still on 2.40.x, update to the latest release so the merged fixes, including PR #30375, are actually running. Self-hosted users can follow our install n8n on Ubuntu VPS guide for the update and restart steps.
Pitfalls that make it look like the fix failed
- Testing with old mail. After the rewrite, Execute step waits for a fresh email: mail already unread before you connected will not fire the trigger. Send a brand-new test message.
- Force Reconnect left at 60. Enabling the option without lowering the value means the first rescan is an hour away. Set it to 5, or 1 while testing.
- Fetch Only New Emails off. Each reconnect then re-emits every unread message, so one fix looks like a flood of duplicates.
- Live testing still hangs. Several users reported that manual Execute step never fires while production runs work, a known limitation of the rewritten trigger. Judge by production executions, not by the editor.
- Another client marking mail read. If Outlook or another mail client marks messages as read before n8n sees them, Fetch Only New Emails has nothing to fetch.
How to verify
Set Force Reconnect Every Minutes to 1, send a new email to the mailbox from another account, and wait about a minute. Then open Executions: a production execution should appear with the new message as its input. If it does, set Force Reconnect back to 5. If nothing appears after two reconnect cycles, check the instance log, because the editor will not show the real error.
When this does not apply
If the trigger never worked, even before any update, the cause is usually the mail server. Many shared hosts accept IMAP IDLE but never send the new-mail notification. Confirm IDLE support with your provider, or move the mailbox to Gmail or Outlook and use the dedicated Gmail or Outlook trigger, which polls on a schedule.
If the workflow deactivates itself the moment you turn it on, that is a different failure: the trigger connection itself is failing at activation. Check the instance log for the real error.
If your workflow shows “Published, has changes” after you edit the trigger node, production still runs the old version until you publish, as our guide to fixing a workflow that stays on the old version after publishing explains. For the full reference, see our complete n8n guide.
Frequently asked questions
Why does my IMAP trigger work when I run it manually but never fire in production?
A manual run performs a one-off search of the mailbox, while the published trigger holds an IMAP connection open and waits for the server to announce new mail. After the 2.40 rewrite, if your server never sends that announcement, the published trigger waits forever with no error, even though manual runs keep working.
Do I have to delete the whole workflow to fix this?
No. Delete only the Email Trigger (IMAP) node and add a fresh one with the same settings. That resets the last-seen UID watermark, which can go stale during an upgrade. The rest of the workflow stays untouched.
Will Force Reconnect every 5 minutes send my emails through twice?
Not if Fetch Only New Emails stays on. Each reconnect then rescans only for unread mail the trigger has not handled yet. Turn that setting off, and every rescan re-emits all unread mail, which looks like duplicates.