Duplicate without user-selected canonical: make the choice yourself

Updated · Check Before Publish

"Duplicate without user-selected canonical" means Google judged this URL a duplicate of another page, you gave no preferred canonical, and Google chose the other page, so it won't serve this one. Google calls this working as intended, not an error. If you disagree with its choice, mark the canonical explicitly, or make the pages' content differ substantially.

How it differs from the status next to it

  • Duplicate without user-selected canonical: you declared nothing. Google chose.
  • Alternate page with proper canonical tag: you did declare a canonical, it points to a page that is indexed, and Google says there is nothing you need to do.
  • Duplicate, Google chose different canonical than user: you declared one and Google thinks another URL makes a better canonical. Google's help page says to compare this page, the canonical you declared and the one Google picked. If the canonical you declared isn't similar to this page, Google won't choose it.

Where the duplicates come from

  • The same page reachable with and without parameters: ?utm_source=, ?ref=, ?sort=
  • Both www and non-www, or both http and https, serving the content instead of redirecting
  • Trailing-slash and non-trailing-slash versions, or mixed upper and lower case
  • Product variants, or the same article under two category paths
  • A staging or preview domain that is publicly reachable

Check what this page declares, and whether it loads

The check below reports whether the page has a canonical at all, whether it points at this URL or another one, whether there are several conflicting canonicals, and it fetches the canonical target to see whether that URL actually returns 200, redirects, or fails. A canonical that points at a URL you didn't intend is a mistake to fix: Google lists incorrect canonical elements among the common reasons it picks a different canonical.

Opens the Meta, Canonical & Noindex Checker and runs it. Free, no account. The address is not stored and never goes into the URL bar.

How to fix it

  1. Decide which URL is the real one. Usually the clean, parameter-free, https, canonical-host version.
  2. Put a self-referencing canonical on it, as an absolute URL, written exactly as the page is served:
    <link rel="canonical" href="https://www.example.com/products/kettle">
  3. Point the duplicates at it with the same tag. If you want to get rid of a duplicate page entirely, redirect it instead: www to non-www (or the other way round), http to https. Google treats URLs that differ by a trailing slash or by upper and lower case as different URLs, so settle on one form.
  4. Make your internal links and sitemap agree. If every link on your site points at the parameter version, your site is sending two different answers about which URL is the main one.
  5. Keep staging out of the index with authentication, and make sure it never carries canonicals pointing at itself in production.

Worth saying plainly: the canonical is a strong hint, not a command. If the pages really are near-identical, Google may still fold them together. The way to stop that is to make the pages genuinely different, or to accept the consolidation.

How to confirm the fix

Run the check again until it reports that the canonical target is this page and returns 200. Then use URL Inspection and compare User-declared canonical with Google-selected canonical. Google says re-evaluation takes time: it might hold pages in a duplicate cluster for up to two weeks after you fix them, and URL Inspection shows data from Google's last crawl of the page.

If the page still does not appear in search after that, work through why a page isn't showing up on Google.

The status definitions and the advice to mark a canonical explicitly, or make the pages' content differ, are Google's. Which URL Google selects is its own decision.

Sources

We read these official pages on 2026-10-13.

Fixed this one? Before you send people to the page, run the full pre-publish check for everything else: indexing signals, link preview, broken links, alt text, structured data and HTTPS in one pass.