How to Add and Check a Canonical Tag in HTML, WordPress and Shopify

What a canonical tag is and what it does

A canonical tag is a <link> element with rel="canonical" placed in the <head> of a page. It names the preferred address for the content on that page. When Google finds several URLs showing the same or nearly the same content, it groups them together, picks one to index and show in results, and folds the signals from the others into it. Your canonical tag is your vote in that choice.

It is a strong hint, not an order. Google also weighs redirects, internal links, the sitemap, hreflang and whether the pages are truly duplicates. If your tag disagrees with everything else, Google may pick a different URL, and Search Console will say so. That is why adding the tag is only half the job; the other half is making the rest of the site agree with it.

There are two kinds of canonical you will use:

• Self-referencing. The page points to its own clean address. This is the normal case and belongs on every page you want indexed. It protects you when someone links to the page with ?utm_source= or other parameters.
• Cross-page. A duplicate or near-duplicate points to a different, main page. Use this for printer versions, sorted or filtered listings, syndicated copies and product variants that do not deserve their own result.

A canonical tag is not a way to merge two different pages, and it does not hide a page from visitors. If an old page should never be visited again, a 301 redirect is the honest answer. The canonical fixes guide covers how canonicals fit alongside redirects and noindex.

Canonical tag code and examples

The tag itself is short. This is a self-referencing canonical for a service page:

<link rel="canonical" href="https://www.example.com/services/wedding-cakes/">

And this is a cross-page canonical on a filtered listing that should consolidate into the main category:

<!-- On https://www.example.com/cakes/?sort=price&flavour=lemon -->
<link rel="canonical" href="https://www.example.com/cakes/">

Five rules make the difference between a tag Google trusts and one it ignores:

Relative addresses are allowed by the HTML specification, but they break easily when a page is copied, proxied or served on a staging host, so absolute addresses are the safer habit. Keep the trailing slash consistent with how your server actually answers, because /cakes and /cakes/ are different URLs.

Canonicals for files that are not HTML

A PDF cannot carry a head element, so you send the canonical as an HTTP response header instead. Google supports this for web search:

Link: <https://www.example.com/guides/cake-care.html>; rel="canonical"

How to add a canonical tag in HTML

On a hand-built or static site, open the template or each page file and add the tag inside the head, near the title and meta description:

<head>
  <meta charset="UTF-8">
  <title>Wedding Cakes in Leeds | Example Bakery</title>
  <meta name="description" content="Tiered wedding cakes baked to order.">
  <link rel="canonical" href="https://www.example.com/services/wedding-cakes/">
</head>

If you use a static site generator or a shared header include, add it once in the layout and build the address from the page's own path, so every page gets its own value. A common mistake is pasting the home page address into a shared header, which tells Google that every page is a copy of the home page. Pages then drop out of the index one by one.

For a server-side template, the pattern is the same: print the site's base address plus the clean path, without query strings. In PHP that might look like this:

<?php
$path = strtok($_SERVER['REQUEST_URI'], '?');
$canonical = 'https://www.example.com' . $path;
?>
<link rel="canonical" href="<?= htmlspecialchars($canonical) ?>">

That strips every parameter, which is right for tracking codes but wrong if a parameter genuinely changes the content, such as ?page=2 on a listing. Keep those parameters in the canonical, so page 2 points to page 2. The pagination guide explains why paginated pages should not all point to page 1.

How to add or fix the canonical URL in WordPress

WordPress core already prints a self-referencing canonical on single posts and pages through its rel_canonical() function. SEO plugins replace that output with their own and extend it to archives, categories and the home page. So on most WordPress sites you are not adding a canonical; you are checking and correcting the one the plugin prints.

Yoast SEO

Edit the post or page, open the Yoast SEO panel, choose the Advanced section and fill in Canonical URL. Leave it empty for a normal self-referencing canonical; only enter an address when this page is a duplicate of another. For template-wide changes, developers can use the wpseo_canonical filter:

add_filter( 'wpseo_canonical', function ( $url ) {
    if ( is_page( 'cake-menu-print' ) ) {
        return home_url( '/cake-menu/' );
    }
    return $url;
} );

Rank Math

Edit the post, open the Rank Math panel, go to the Advanced tab and use the Canonical URL field. The matching developer filter is rank_math/frontend/canonical.

Common WordPress canonical faults

• Two canonical tags. A theme hard-codes one in header.php while the plugin prints another. Remove the theme copy.
• Two SEO plugins active. Yoast and Rank Math together, or an SEO plugin plus an all-in-one theme framework, often output duplicate tags. Keep one.
• Wrong host. The WordPress Address and Site Address settings still say http or the old domain after a move, so every canonical points to the wrong place. See the WordPress migration guide.
• Cached HTML. A page cache keeps serving the old tag after you change it. Purge the cache before you test.

How to fix the canonical URL in Shopify

Shopify themes print the canonical from the Liquid canonical_url object, normally in layout/theme.liquid. Open Online Store, then Themes, then Edit code, and search the layout for canonical. You should find one line like this:

<link rel="canonical" href="{{ canonical_url }}">

Shopify handles the common duplicates for you. A product reached through a collection, such as /collections/cakes/products/lemon-drizzle, gets a canonical to /products/lemon-drizzle. Leave that behaviour alone; it is correct.

The faults that do happen on Shopify are usually these:

• Missing tag. An older or heavily edited theme lost the line. Add it back inside the head of theme.liquid.
• Duplicate tag. An SEO app injects a second canonical. Check the rendered HTML and turn off the app's canonical option, or remove the theme line, so only one remains.
• Hard-coded address. A developer replaced {{ canonical_url }} with a fixed domain, so every page claims to be the home page.
• Wrong primary domain. If the store's primary domain in Settings, then Domains, is not the one you want indexed, canonicals follow it. Set the primary domain first.

Product and collection duplicates go deeper than the tag; the duplicate products guide covers variants, near-identical products and filtered collections.

How to check which canonical a page really outputs

The settings screen tells you what you asked for. The served HTML tells you what Google receives. Always check the second. From a terminal, fetch the page and pull out the tag:

curl -sL https://www.example.com/services/wedding-cakes/ | grep -io '<link[^>]*canonical[^>]*>'

More than one line in the result means more than one canonical. To check for a canonical sent as an HTTP header, look at the response headers:

curl -sI https://www.example.com/files/cake-care.pdf | grep -i '^link:'

In the browser, open Chrome DevTools, go to the Console and run:

[...document.querySelectorAll('link[rel="canonical"]')].map(l => l.href)

This shows the rendered page, after scripts have run. If curl shows one value and DevTools shows another, a script is rewriting the tag, and you should fix it at the source.

What Google decided

In Google Search Console, paste the address into URL Inspection and open the page indexing details. You will see the User-declared canonical, which is your tag, and the Google-selected canonical. When they match, you are done. When they differ, Google has overruled you, usually because internal links, the sitemap or a redirect point elsewhere, or because the content is a near-duplicate of another page. The canonical conflicts guide walks through that case.

Checking the whole website

To check a website for canonical tags at scale, run a crawl with the site crawler and sort pages by canonical status: missing, self-referencing, pointing elsewhere, pointing to a redirect or a 404, and pages with more than one tag. That list is your work queue.

The fix, step by step

Edge cases worth knowing

• Canonical plus noindex. Pointing a page's canonical at another URL while also marking it noindex sends mixed signals. Pick one: canonical for a duplicate you want consolidated, noindex for a page that should not appear at all.
• Canonical to a different language. Each language version should canonicalise to itself and link to the others with hreflang. A French page canonicalising to the English one drops out of French results. See hreflang and canonical conflicts.
• Canonical chains. Page A points to B, and B points to C. Google may follow it, but point A directly at C.
• www and non-www. The canonical should use the same host your redirects settle on. Settle that first with the www canonicalisation guide.
• Very different content. If the two pages are not really duplicates, Google tends to ignore the tag. A canonical cannot make an unrelated page inherit another page's place in results.
• Syndicated articles. If a partner republishes your article, ask them to noindex their copy; Google no longer recommends cross-domain canonicals for syndication, and a link back to the original helps too.

Keeping canonicals correct

Canonicals break quietly. A theme update, a new plugin, a domain move or a caching layer can change every tag on the site at once, and nothing looks wrong to visitors. Build three habits:

1. Check after every template change. After a theme update, a plugin install or a new page builder, run the curl check on one page of each template: home, category, product or post, and a static page.
2. Crawl monthly. A scheduled crawl catches missing, duplicated and wrong canonicals before Google acts on them. Compare the count of self-referencing pages against last month.
3. Keep signals aligned. Your sitemap should list only canonical addresses, menus should link to canonical addresses, and redirects should end on canonical addresses. When all four agree, Google almost always follows your tag.