Premium tools, free for everyone. Most run right in your browser.

OpenMultiURL

SEO

How to Check Canonical URLs and Verify Google's Choice

How to Check Canonical URLs and Verify Google's Choice title on a coloured card

A page can declare one canonical URL while Google chooses another. That's the problem I want you to catch before you call a canonical check finished. When I explain how to check canonical URLs, I separate what the page declares, where that address leads, and what Google has indexed. I built Open Multiple URLs and test my tools on a budget Android phone and a laptop. But opening pages is only the start, not proof that their canonical setup is right.

What am I checking when I check a canonical URL?

I think of a canonical as a preferred address for duplicate or very similar content. A tracking URL and its clean version might show the same article. Rather than treating the tag as an order, Google considers it alongside other signals when choosing which URL represents that content.

The declared URL and Google's choice are different checks

The user-declared canonical is the preference your site supplies. The Google-selected canonical is the address Google chose during indexing. I don't use those names as if they mean the same thing. A browser can help me inspect a declaration, but it can't tell me Google's indexed decision just by loading a page.

  • I inspect the declaration: what canonical does the page actually supply?
  • I inspect the destination: is it the right, accessible page?
  • I inspect Google's decision: does indexed URL Inspection report the same address?

Google's canonicalization overview explains that canonical declarations are hints, not rules. So I don't promise that adding a tag will force indexing or remove another URL from search. I want a consistent preference supported by the site's actual behaviour, not just a tag that looks tidy.

How do I find the canonical URL in page source?

I start with the exact address under review, including its parameters. I keep that address in my notes before opening it, because a redirect might change the browser's address bar. If I inspect only the destination, I can miss what happened to the original URL.

Read the original HTML first

  1. I open the URL on my laptop and note the address that finishes loading.
  2. I right-click the page and choose View page source, where my browser provides that option.
  3. I search for canonical and find the link element whose rel value is canonical.
  4. I copy its href value and check that the element sits inside the document head.
  5. I look for another canonical declaration rather than stopping at the first match.

For an illustrative article, I might expect the href value https://example.com/article/. I want the complete address, including https:// and the intended hostname. Google supports relative canonical paths, but recommends absolute URLs. I prefer the full address because I can compare it directly without working out which base URL a relative path depends on.

If I find nothing, I record 'not found in source', not 'no canonical anywhere'. JavaScript or a response header may still supply one. A missing declaration also doesn't mean Google cannot choose a canonical. It means I haven't found an explicit preference in that particular place.

How do I check rendered canonicals and HTTP headers?

Compare source with the rendered page

Next, I right-click the loaded page and choose Inspect. In the Elements panel, I search for canonical and compare the href with the original source. This view shows the current document after browser-side changes. It can reveal a canonical inserted or changed by JavaScript, but it still isn't a report of Google's indexed choice.

If the addresses disagree, I treat that as something to fix, not a reason to pick whichever result I prefer. Google's JavaScript SEO guidance says JavaScript shouldn't change the canonical to a different value from the original HTML. I want the server output and rendered document to agree.

Look at response headers for PDFs too

A PDF doesn't have an HTML head to inspect. I open the browser's Network panel, reload the file, select its request, and check Response Headers for a Link header declaring the canonical relationship. Google supports this method for supported documents, including PDFs. I also check headers when an HTML page's declarations seem incomplete or conflicting.

Tip from me: I label every result by where I found it: original HTML, rendered document, HTTP header, or Google indexed data. That small note prevents me from comparing different kinds of evidence as though they were identical.

How do I test whether the canonical target is correct?

I open the declared target separately and compare its main content with the starting page. Similar titles aren't enough. If an article points to an unrelated category, or every product points to the homepage, the declaration doesn't describe duplicate content. I settle that question before spending time on smaller URL differences.

My checks for a canonical destination
CheckWhat I want to seeWhat I investigate
AddressAn absolute URL for the intended pageWrong host, protocol, path, or parameters
ResponseA successful page response, usually 200Errors or an unexpected redirect
IndexabilityA target without noindexA target that asks not to be indexed
ContentDuplicate or very similar primary contentA substantially different page
Supporting signalsLinks and sitemap using the preferred addressDifferent preferred URLs across the site

I use Redirect Checker alongside the browser to review the destination and redirect behaviour. A target that redirects deserves another look: I normally want the canonical to name the final preferred page directly. A successful response is a useful diagnostic check, though, not a guarantee that Google will select that URL.

I also check whether crawling is restricted and whether the target carries noindex. Those aren't interchangeable settings, so I record them separately. Then I compare navigation links and sitemap entries with my intended URL. Google's canonical methods guidance recommends self-referencing canonicals on canonical pages, so I inspect the destination's own declaration too.

How do I use Open Multiple URLs for canonical checks?

I use Open Multiple URLs to put related pages within reach for a manual comparison. It opens addresses; it doesn't extract canonical tags, inspect headers, or report Google's choice. I've tried this workflow with Open Multiple URLs, and I keep the opening step separate from the actual inspection.

  1. I collect the exact URLs for one suspected duplicate group and paste them into Open Multiple URLs.
    • I include the preferred page and the parameter or host variants I need to test.
    • I keep the original list so redirects don't erase the starting addresses from my notes.
  2. I open the group and compare each tab's final address with the original entry.
  3. I inspect each page's source, then its rendered document or response headers where needed.
  4. I record the declared canonical and open its destination if it isn't already in the group.
  5. I finish with indexed URL Inspection for pages whose Google-selected canonical I need to confirm.

For tracking variants, I can use Tracking Parameter Remover to prepare a clean candidate for comparison. I don't treat its output as proof of the right canonical. First I check that removing the parameters preserves the same primary content. A content-changing filter is a different question from a campaign tag.

Tip from me: On my budget Android phone, I keep comparison groups small enough to follow without losing my place. If opening tabs is blocked, I check the browser's pop-up notice. For source and header work, I prefer my laptop.

How do I check Google's selected canonical in Search Console?

Once I know what the site declares, I open the relevant property in Google Search Console. I inspect the exact URL, rather than a cleaner address that I assume represents it. My question is now different: what did Google select for this URL in its indexed data?

  1. I enter the address into URL Inspection.
  2. I review the indexed result and expand the Page indexing details.
  3. I compare User-declared canonical with Google-selected canonical where those fields are available.
  4. I note the indexed information's timing before comparing it with today's live page.

The live test doesn't predict the canonical choice

I can use Test live URL to investigate the current page, but I don't present that result as confirmation of Google's canonical selection. The URL Inspection documentation makes the distinction clear: the live test cannot predict whether Google will consider a URL canonical. Indexed results describe the most recently indexed version, which may differ from today's page.

If the values disagree, I revisit the evidence rather than repeatedly changing the tag. I compare content, redirects, HTTPS, host choices, sitemap entries, and internal links. For regional versions, I review the canonical and hreflang setup together. A mismatch is a reason to investigate, not automatic proof that the tag's syntax is broken.

Which canonical should I use for common URL variants?

I decide whether two URLs are duplicates before deciding whether they should share a canonical. A shared template or topic isn't enough. This matters most with filters, product variants, regional pages, and pagination, where a tidy-looking rule can collapse pages that actually show different content.

How I choose between a canonical, a separate page, and a redirect
SituationMy starting decisionWhat I verify
Tracking parameter, unchanged contentConsider the clean URL as canonicalThe primary content really matches
Duplicate URL that must remain accessibleConsider a canonical to the preferred versionBoth pages are genuinely equivalent
Retired or moved duplicatePrefer a server-side redirectThe destination is the intended replacement
Filtered page with distinct contentDon't automatically canonicalize to the categoryIts content and intended search purpose
Later pagination pageGenerally use its own canonicalIt isn't a duplicate of the first page
Regional version in the same languageReview canonical and hreflang togetherRegional pages aren't collapsed indiscriminately

For pagination, I follow Google's pagination guidance: later pages generally need distinct canonical URLs when they aren't duplicates of the first page. A listing with different items isn't the same page just because its heading and layout match.

For a retired address, I prefer a redirect because the old page no longer needs to stay available separately. Google recommends redirects for retiring duplicate URLs. I don't describe a canonical on a redirecting response as if visitors also load a separate HTML page there; I inspect the redirect and the final destination as separate parts of the journey.

What canonical mistakes should I avoid?

My first fix is usually consistency, not another declaration. Adding a second canonical can make the problem harder to read. Google's guidance on common canonical mistakes warns about multiple declarations, missing targets, and targets marked noindex.

  • I don't point unrelated pages at the homepage just to reduce the number of indexed URLs.
  • I don't leave competing canonical values in the source, rendered page, or response headers.
  • I don't point at a missing destination, an unintended redirect, or a noindex page.
  • I don't use a canonical as a reliable way to keep a page out of search.
  • I don't assume every parameter can be removed without changing the content.

When an address looks almost right, I use URL Parser to compare its parts. A different hostname, protocol, or path can be easy to overlook in a long URL. Parsing helps me see the difference; it doesn't establish that the page exists or that Google chose it.

I also avoid calling 'Alternate page with proper canonical tag' an error by default. If I intended that URL to be a duplicate of another page, the status can describe the expected outcome. I investigate when the selected representative is wrong, or when a page I wanted treated separately has been grouped with something else.

How do I check canonical URLs across a whole website?

For a website-wide check, I start with a URL export from a crawler, sitemap, or CMS. I want the original address, final destination, response status, declared canonical, and indexability observations in separate columns. If a crawler renders JavaScript, I record that too, rather than assuming every canonical checker reads the same version.

  • I group URLs by template, directory, parameter pattern, and declared canonical destination.
  • I flag missing declarations, conflicting values, unexpected hosts, and targets that redirect or fail.
  • I manually inspect representative pages from each pattern before changing a shared template.
  • I use indexed Search Console data for the Google-selection question, keeping it separate from fetched-page results.

I can use URL Deduplicator to prepare a review list, but I preserve meaningful variants for testing. If I remove every parameterized or alternate-host address before the audit, I may remove the exact examples that expose the problem. Cleaning the list and investigating duplicate content are different jobs.

Before trusting any report, I ask what it actually checked: raw HTML, rendered HTML, headers, or Google's indexed data. A third-party checker can report what it fetched. It generally cannot establish Google's chosen canonical from that fetch alone. My finished report keeps those limits visible and gives you a specific change to test, rather than a vague 'canonical failed' label.

Questions people ask

How do I check a page's canonical URL?

I inspect the original HTML for a canonical link in the head, copy its href, and open that target. Then I compare rendered output or headers where needed and review redirects, content, and indexability. To learn whether Google selected that address, I check indexed URL Inspection data separately.

Does View Source or Inspect show the canonical Google sees?

I use View Source to read the original HTML and Inspect to read the browser's current document after changes. Neither alone proves what Google indexed or selected. If JavaScript changes the canonical, I compare both versions and fix conflicting values rather than assuming the rendered one must win.

Why does Google choose a different canonical than the one I declared?

I first remember that the declaration is a hint. Google considers signals gathered during indexing, including redirects, HTTPS, sitemap inclusion, and canonical annotations. I check those alongside content similarity and internal linking. I also compare dates, because indexed URL Inspection may describe an older version than the live page.

Should a canonical tag point to the page itself?

For the preferred canonical page, I follow Google's recommendation to use a self-referencing canonical. A duplicate page may instead declare the preferred equivalent. I don't make every page point to itself without considering duplicates, and I don't make distinct pages share one target just because their topics overlap.

Can I use a canonical tag on a PDF or other non-HTML file?

For a PDF, I inspect the HTTP response headers rather than looking for an HTML head. Google supports canonical information through a Link response header for supported documents, including PDFs. I still check the destination and surrounding signals, because using a header doesn't turn the canonical hint into a command.

Should paginated pages all canonicalize to the first page?

I generally don't point every pagination page at the first one. Google's guidance recommends distinct canonical URLs for pagination pages that aren't duplicates. If later pages show different items or content, I preserve that distinction rather than treating a shared layout as proof that the pages are equivalent.

What to do next

I would start with one duplicate group, not a site-wide template edit. Save the original addresses, find each declaration, compare source with rendered output where needed, and test the target. Then check indexed URL Inspection to see whether Google's choice matches your preference. If it doesn't, review the content and supporting signals before changing anything. You can use Open Multiple URLs to open the group for that comparison. Keep your notes split between what the site declares, what the destination does, and what Google indexed. That's how I turn a quick tag lookup into a useful canonical check.

Sources I checked

Tools from this guide

Keep reading