Updated
“Duplicate without user-selected canonical” means Google found two or more URLs with the same content, none of them declared a canonical, and Google chose another URL to index instead of this one. This URL won’t appear in search. Fix it when Google picked the wrong version or when the duplicate shouldn’t exist: declare a canonical, redirect the duplicate, or make the pages genuinely different.
What it means
Google’s Page indexing help describes this URL as a duplicate of another page that “doesn’t indicate a preferred canonical page,” so Google chose the other page and won’t serve this one in search. It adds that this is not an error but working as intended, because Google doesn’t serve duplicate pages.
The key word is without. Nobody told Google which version is the main one, so Google decided. Most of the time it decides well. You care when it picked a URL you don’t want people to land on (the http:// version, a URL with ?sort= on it) or when the “duplicate” is a page you expected to rank on its own.
The three canonical statuses compared
| Alternate page with proper canonical tag | Duplicate without user-selected canonical | Duplicate, Google chose different canonical than user | |
|---|---|---|---|
| Did this URL declare a canonical? | Yes, pointing to another URL | No | Yes (often itself) |
| Did Google agree? | Yes | Nothing to agree with; Google picked one | No; Google picked a different URL |
| Which URL is indexed | The canonical you declared | The URL Google chose | The URL Google chose |
| Usually means | Working as intended | Duplicates with no canonical set | Your signals disagree, or the pages are not true duplicates |
| Action | Usually none; check the target is right | Pick the canonical and declare it, or redirect | Align your signals, or accept Google’s choice |
| Guide | Alternate page with proper canonical tag | This page | Google chose different canonical |
The confusion between this status and Duplicate, Google chose different canonical than user is common, and the fix differs. Here you never declared a canonical. There, you declared one and Google overruled it.
See which URL Google picked
- Open Indexing > Pages in Search Console and click Duplicate without user-selected canonical.
- Click the inspect icon next to an example URL.
- Under Page indexing, User-declared canonical will say None, and Google-selected canonical shows the URL Google indexed instead. Open both in your browser and compare them.
Do this for five or ten examples. The pattern behind them (all trailing slashes, all one filter, all one listing reached by two paths) is usually obvious by then. Note that the live test doesn’t check canonical selection, so rely on the indexed result for this (URL Inspection help). If Google picked a URL on a host you haven’t verified in Search Console, you won’t see that URL’s traffic, which is another reason to settle the canonical yourself.
Causes, with how to fix each
Google’s canonicalization guide lists the usual sources of duplicates: region variants, mobile and desktop versions, http and https, sorting and filtering, and accidental copies such as a demo site left open to crawlers. On a directory they look like this:
| Cause | Example | Fix |
|---|---|---|
| Protocol and host variants | http://example.com/plumbers and https://www.example.com/plumbers | Site-wide 301 redirect to one protocol and host. |
| Trailing slash or letter case | /listing/joes-plumbing/ and /listing/Joes-Plumbing | Redirect to one form; link only to that form. |
| One listing under several category paths | /plumbers/austin-tx/joes-plumbing, /water-heaters/austin-tx/joes-plumbing | One canonical listing URL; every path links to it, or redirects to it. |
| Sort, filter, view and page-one parameters | ?sort=rating, ?view=map, ?page=1 | Canonical tag to the clean URL; link to the clean URL; redirect ?page=1 to the page itself. |
| Print and share versions | /listing/joes-plumbing?print=1 | Canonical to the listing, or don’t make print views crawlable. |
| Tracking parameters | ?utm_source=newsletter, ?ref=partner | Canonical to the clean URL. |
| Empty or near-empty template pages | /plumbers/hutto-tx and /plumbers/manor-tx, both “No plumbers found” | Fill them, noindex them, or don’t publish them. Empty pages can also turn up as soft 404s. |
| Staging or mirror hosts | staging.example.com copying the live site | Password-protect staging. A login keeps it out of Google; robots.txt alone doesn’t. |
How to fix duplicate without user-selected canonical
- Decide which URL should be the canonical for each group: the clean, https, preferred-host version that your navigation links to.
- If the duplicate doesn’t need to exist, redirect it with a 301. Protocol, host, trailing-slash and old-path variants all fall here. A redirect is the strongest canonical signal there is.
- If the duplicate must exist (a sort order, a filter, a print view), add
<link rel="canonical" href="…">pointing to the main URL, and put a self-referencing canonical on the main URL itself. - Point internal links and sitemaps at the canonical URL only. Linking to
/listing/joes-plumbing/in one template and/listing/joes-plumbingin another is how many of these duplicates start. - If the pages are not really duplicates, such as two city pages with the same boilerplate and no listings, make them substantially different. Google’s help gives exactly this advice when you think a page isn’t a duplicate of the one Google picked.
- Then give it time. Google’s canonical troubleshooting guide says pages may stay in a duplicate cluster for up to two weeks after a fix, and split out faster when the difference is clear. Request indexing for the few most important URLs.
Two things not to do, both from Google’s canonical guide: don’t use noindex to choose a canonical within your site (it removes the page from search entirely), and don’t use robots.txt for canonicalization (Google may still index blocked URLs, without their content).
Check that the fix worked
After Google recrawls the URLs, their status should change, and the new status tells you which fix took hold:
- Added a canonical tag: the duplicate should move to Alternate page with proper canonical tag. That is the result you want.
- Added a redirect: the duplicate should move to Page with redirect.
- Made the pages different: the URL should be indexed in its own right, or move to Crawled - currently not indexed if Google still finds it too thin to keep.
- It moved to Duplicate, Google chose different canonical than user: Google saw your tag but disagrees. Your other signals probably still point at the other URL.
Once the examples you inspected have moved, click Validate fix on the reason in the Page indexing report so Google rechecks the rest.
Does duplicate content hurt your site?
Not in the way people fear. Google’s canonicalization guide says some duplicate content on a site is normal and isn’t a violation of its spam policies. Google simply shows one version. The real costs are practical: Google may show the version you like least, it spends crawling on copies, and links can be split across versions until the duplicates are consolidated.
Directories have one extra risk. Thousands of near-identical city or category pages that exist only to catch searches are a quality problem, not a canonical one, and canonical tags won’t solve it. See doorway pages for the test Google applies.
When to leave it alone
If Google picked the URL you would have picked, and the duplicate is a variant nobody links to, the status is harmless. Many sites carry thousands of these from tracking links shared on social media. Add canonical tags to the templates that produce them when you get to it, so future duplicates are labelled for Google, and move on.
Fixing it at scale on a directory
Directories make duplicates through templates, so fix the template, not the URLs:
- Give every template a self-referencing canonical built from the clean URL, with parameters stripped.
- Generate one canonical URL per listing, and make category, city and search pages link to that URL rather than a path of their own.
- Keep the canonical form consistent everywhere: sitemaps, breadcrumbs, schema
urlfields and internal links. - Don’t publish category-in-city pages until they have listings, or noindex them below a minimum. Empty pages are duplicates of each other.
Then check a sample of each page type, not just the 1,000 examples Search Console shows. The bulk index checker guide explains how.
See duplicates by page type in DirectoryLens
In DirectoryLens’s Google indexing report, the Individual pages tab shows Google’s coverage state for each inspected URL, and each URL’s history shows Google’s canonical next to yours. For this status yours shows as “–”, which makes undeclared duplicates easy to spot within a page type. Start an inspection run on one page type, say every listing URL in your listings sitemap, and the results show how many are duplicates and which URLs Google kept instead, beyond Search Console’s 1,000 examples. Each URL keeps a history of state changes, so you can see when your fix took hold.
The Google picked another canonical flag only fires when you declared a canonical and Google chose a different one. For undeclared duplicates like these, filter by state and read Google’s status on each row. All statuses are explained in the page indexing issues guide.