When an Elementor Loop shows posts outside its selected category, tag or custom taxonomy, first compare the expected post IDs with the actual result. Elementor Pro 4.3.2, released October 13, 2026, includes a fix for term filters being ignored on some sites. Updating is relevant, but incorrect query settings and cached pages can produce similar symptoms.
The vendor changelog identifies the release change. The following controlled content test helps distinguish that regression from a configuration error without deleting your existing loop.
Create a Small, Unambiguous Content Set
On staging, choose three published test items with distinct titles. Assign item A to the target term only, item B to an unrelated term only, and item C to both. Write their IDs in your test notes. Avoid draft or private items for this initial test, because visibility rules would add another variable.
Select the taxonomy actually attached to the loop's post type. A product category, a post category and a custom portfolio taxonomy are not interchangeable merely because they display the same label. Record the term ID as well as its name.
Configure a simple loop with no alternate template, custom query callback or extra exclusion. Define the expected result before loading it. With a plain inclusion of the target term, A and C should qualify and B should not. If your query adds other constraints, document them rather than borrowing this expectation unchanged.
Compare Editor, Public Page and Pagination
| Surface |
Evidence |
| Editor preview |
IDs or permalinks of visible items |
| Logged-out first page |
Whether the same term restriction applies |
| Next page |
No unrelated items or duplicate IDs |
| Empty selection test |
Intended empty state, not an unfiltered catalog |
Use identical sorting for every comparison. A newest-first list with several posts sharing a timestamp can appear unstable, so use a deterministic fixture and record the sort rule. Screenshots of similar-looking cards are less reliable than a list of permalinks.
Query: Correct post type and taxonomy. Identity: Compare IDs, not similar cards. Pagination: No unrelated or repeated items. Public cache: Matches current query output. Explanatory checklist, not a customer test result.
Test the Supported Update
Back up the site, verify the restore path and update a representative staging installation to a supported Core/Pro combination that includes the fix. Repeat the original query before adding customizations back. If the simple loop works but the full page fails, reintroduce custom query logic and extensions individually.
Clear the affected public page cache after changing the query. A correct editor preview does not prove the CDN is serving the current result. Conversely, a cache purge cannot repair a query that is already wrong before caching.
If the filter remains ignored, collect the post type, taxonomy and term IDs, loop settings, current versions and the returned IDs. Share only synthetic content or public permalinks. Do not export your entire site database to demonstrate three incorrect cards.
Should You Hide the Wrong Cards With CSS?
No. Hidden cards can still affect pagination, links and accessibility. Fix the query or rendering path so the returned content matches the intended collection.
Once filtering works, run the larger Atomic Loops regression checklist for alternates and empty states. Get help with a broken dynamic listing when the issue appears only with your site's custom query or caching setup.
References reviewed October 15, 2026. Examples are explanatory, not customer test results.