Canonical Points to a Different Internal URL: When to Fix It

TL;DR: A canonical that points from Page A to a different internal Page B hands Page A's indexing and ranking signals to Page B. That is exactly what you want when the two pages are duplicates and you chose the winner on purpose. It is a slow bleed when a unique, money-earning page is quietly nominating some other URL as its master copy because a template filled the tag in wrong.

What this flag actually means

The check fires when a page carries <link rel="canonical"> whose href is a different URL on the same site than the one you are looking at. In canonical language, the page you are on is "canonicalized" to another address. You are telling Google: this URL and that URL are the same thing, index the other one, pass the link equity there, drop me from the results. Screaming Frog labels these rows "Canonicalised"; the question the audit forces you to answer is whether a human meant to do that or a plugin did it by accident.

A real failing tag, and the fix

Here is a product page that should rank for its own variant but is throwing its weight at a category page instead:

<!-- On https://example.com/boots/chelsea-tan/ -->
<link rel="canonical" href="https://example.com/boots/" />

Every visit, every link, every scrap of relevance for "chelsea tan boots" is being routed to the generic /boots/ listing. The tan Chelsea page will fall out of the index and the category page will not rank for the specific term either. The fix is a self-referencing canonical:

<!-- On https://example.com/boots/chelsea-tan/ -->
<link rel="canonical" href="https://example.com/boots/chelsea-tan/" />

If the cross-canonical was intended (say two near-identical URLs and you picked one), leave it — but confirm the chosen target returns 200 and canonicalizes to itself, or you have just built a chain. More on that in Canonical Points to Another Canonicalized URL.

One more trap worth naming: canonicalizing a paginated ?page=2 back to page 1. That is not consolidation — pages 2, 3 and beyond hold different products or comments, so canonicalizing them to page 1 buries content Google would otherwise crawl and index. Each paginated URL should self-reference. The audit does not know your intent, which is why it surfaces every internal cross-canonical for a human decision rather than auto-classifying them as good or bad.

Intended consolidation vs. accidental self-sabotage

How to detect it

1. Screaming Frog: crawl the site, open the Canonicals tab, and filter to Canonicalised. Add the Indexability column — anything marked "Non-Indexable: Canonicalised" whose URL you expected to rank is a suspect. Sort by Indexability Status to batch them.
2. GSC URL Inspection: paste the page URL, expand Page indexing, and compare "User-declared canonical" with "Google-selected canonical." If the user-declared value is a different internal URL than the one you inspected, that is the flag confirmed straight from Google's mouth.
3. curl: pull the raw head fast — curl -s https://example.com/boots/chelsea-tan/ | grep -i canonical. If the printed href is not the URL you requested, you have the mismatch without opening a browser.

How to fix it

1. Decide intent per URL: is this page a genuine duplicate of the target, or a unique page? Only unique pages need repair.
2. For unique pages, set a self-referencing absolute canonical (full https://, trailing slash matching your site's convention).
3. Trace the bug to its source — usually one line in a theme template or an SEO plugin's canonical override, not 400 hand-edited pages. Fix the function, not the symptoms.
4. Re-crawl to confirm the "Canonicalised" rows dropped, then request indexing in GSC for the highest-value URLs to speed re-inclusion.
5. Sanity-check the targets that legitimately receive consolidation: each should return 200 and canonicalize to itself. See URLs with Duplicate Content: How to Consolidate Them for choosing winners deliberately.

FAQ

Is a canonical to a different internal URL always a bug?

No. It is the correct, deliberate way to consolidate duplicates — parameter URLs, print versions, syndicated variants on your own domain. It is only a problem when a page that deserves to rank on its own is pointing away from itself.

Will the canonicalized page still get crawled?

Yes, Google still crawls it; it just usually will not index it separately and folds its signals into the target. That is why a wrongly canonicalized page vanishes from results even though it returns a clean 200.

Does the canonical target have to be self-referencing?

It should be. If your target is itself canonicalized to a third URL, you have a chain, and Google may ignore the whole thing. Point directly at the final self-referencing 200 URL every time.

Can I use a relative URL in the canonical?

You can, but do not. Relative canonicals break the moment they are rendered on a staging path, a parameterized URL, or a proxied domain. Always use absolute URLs.

How is this different from a redirect?

A 301 sends users and bots to a new URL; a canonical leaves the page live and reachable but tells search engines to credit a different one. If you want the old URL gone entirely, redirect it. If you want it accessible but not indexed separately, canonicalize it. Read more in the Canonical Tags complete reference.

Related canonical checks: Canonical Points to Homepage · Canonical Loop · Duplicate URLs (Technical)