Canonical Tags Explained: When and How to Use rel=canonical

A canonical tag is an HTML link element in the page's <head> that tells search engines which URL is the preferred version of a page. Despite the common name "canonical meta tag", it is not a meta tag. Google treats it as a strong hint, not a command. It looks like this:

<head>
  <title>Blue widgets</title>
  <link rel="canonical" href="https://www.example.com/products/blue-widget/">
</head>

Any duplicate URL, such as https://www.example.com/products/blue-widget/?utm_source=newsletter, carries the same tag, so search engines know which address to index.

Most websites serve the same content at more than one URL without meaning to. Tracking parameters, http and https, trailing slashes, and product pages that live under several categories all do it. A canonical tag tells search engines which version you want indexed. Used badly, it can quietly push your best pages out of Google. This guide explains how canonicals work, when to choose them over a redirect, and how to fix the conflicts that come up most in audits.

What canonicalization means

When Google finds several pages with the same or very similar main content, it groups them and picks one canonical URL to show in search results. The others are treated as duplicates. Google crawls them less often and they rarely appear in results.

Google makes that choice from several signals:

• Redirects. A strong signal that the redirect target should be canonical.
• rel="canonical" annotations. Also a strong signal.
• Sitemap inclusion. A weak signal.
• HTTPS over HTTP. HTTPS URLs are preferred.
• Internal links and how consistently you use one version.

Google calls your declared canonical "a hint, not a rule." When your signals disagree, Google decides for you.

rel=canonical syntax: the link element

Put it in the <head> of every indexable page:

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

Each part has a job:

You will see it written as "rel canonical", "rel=canonical", "link rel canonical" or "canonical link". They all mean this one element. A trailing slash before the closing bracket (/>) is optional in HTML and makes no difference.

Is it a meta tag? Meta canonical vs link rel canonical

People often search for a "meta canonical" or "canonical meta tag", but the canonical is a link element, not a meta element. Google's documentation describes it as "a rel="canonical" link element" and lists the supported ways to declare a canonical: this link element, a Link HTTP header, and sitemap inclusion. Something like <meta name="canonical" content="..."> is not one of those documented methods, so don't rely on it. If an audit tool or plugin calls your canonical a "meta tag", check the HTML: it should start with <link rel="canonical".

Rules that matter:

• Use absolute URLs. Google recommends https://www.example.com/page/ over /page/. Relative canonicals break when a page is served on a different host, like a staging domain or a CDN hostname.
• Only the <head> counts. Google says the element "is only accepted if it appears in the <head> section of the HTML". A canonical that ends up in <body>, for example because a script injected content above it or the HTML is malformed, is ignored.
• One canonical per page. Two different canonical tags is one of the most common CMS bugs. It usually happens when a theme and an SEO plugin both output one. Google may then ignore both.
• No fragments. https://example.com/page#section isn't a valid canonical.

Self-referencing canonicals

Every indexable page should point its canonical at itself. This protects you against parameter variants that other sites or ad platforms create, like ?utm_source=newsletter, ?fbclid=... or ?gclid=.... All of them consolidate back to the clean URL.

Canonicals for PDFs and other files

Non-HTML files can't hold a <link> tag, so send an HTTP header instead:

Link: <https://www.example.com/guides/pricing-guide.pdf>; rel="canonical"

On Apache:

<Files "pricing-guide-print.pdf">
  Header add Link "<https://www.example.com/guides/pricing-guide.pdf>; rel=\"canonical\""
</Files>

Pick one method per URL, either the HTML tag or the header. Google says using both is more error-prone.

Canonical tag vs canonical redirect (301)

Both tell Google which URL you prefer, but they work differently:

• A canonical tag leaves both URLs working. Visitors can still open the duplicate, and search engines are asked to index the canonical instead.
• A 301 redirect (often called a canonical redirect) sends visitors and crawlers from the old URL to the new one, so the duplicate effectively stops existing as a page.

Google's documentation rates both as strong signals: a redirect is "a strong signal that the target of the redirect should become canonical", and rel="canonical" is "a strong signal that the specified URL should become canonical". The difference is purpose. Google says to use redirects "when you want to get rid of existing duplicate pages", and only when deprecating a duplicate.

A redirect is the more predictable choice and the better experience, because the user lands on the right URL. Use a canonical when the duplicate needs to stay reachable. A site-wide canonical redirect usually sends every host and protocol variant to one preferred version. On Apache, for example:

RewriteEngine On
RewriteCond %{HTTPS} off [OR]
RewriteCond %{HTTP_HOST} ^example\.com$ [NC]
RewriteRule ^ https://www.example.com%{REQUEST_URI} [L,R=301]

The redirect target and the canonical tag should always agree. If http://example.com/page/ redirects to https://www.example.com/page/, that final URL must also be the one in the page's canonical tag and your sitemap. Our redirect checker shows each hop and the final status code, and the guide to 301 vs 302 redirects covers which redirect type to use.

Common canonical mistakes and how to fix them

1. Canonical plus noindex on the same page

<!-- Conflicting signals -->
<meta name="robots" content="noindex">
<link rel="canonical" href="https://www.example.com/shoes/">

The canonical says "this page is a duplicate, pass its signals to /shoes/." The noindex says "drop this page." Google advises against using noindex to choose a canonical within a site, because the noindex can end up applied to the target. Pick one:

• If it's a duplicate, keep only the canonical.
• If it has no value at all, keep only the noindex, and remove the canonical or make it self-referencing.

2. Canonical pointing to a redirect

Page A's canonical points to B, and B 301-redirects to C. Google has to work out what you meant. Always point canonicals at the final 200 URL. This often breaks after an http-to-https or domain migration, when templates still print the old host.

3. Canonical pointing to a 404 or noindexed page

If the target is broken or noindexed, Google usually ignores the canonical, and it may index the duplicate or nothing at all. Fix the target, or point to a live, indexable page.

4. Canonical chains

A canonicals to B, which canonicals to C. Google may follow it, but don't rely on that. Every duplicate should point directly at the final canonical.

5. Sitemap and canonical disagree

Your sitemap lists https://example.com/product?id=42, while the page's canonical says https://example.com/products/blue-widget/. Google's documentation specifically calls out mixing canonicals across methods as a mistake. List only canonical URLs in your XML sitemap.

6. Internal links pointing at the non-canonical version

If your navigation links to /Shoes/?sort=popular while the canonical is /shoes/, you're sending mixed signals. Update internal links to point at canonical URLs.

7. Duplicate blocked by robots.txt

If the duplicate URL is disallowed, Google can't crawl it to see the canonical tag. Leave duplicates crawlable. Google says not to use robots.txt for canonicalization. See the robots.txt guide.

8. Paginated pages canonicalized to page 1

Pointing /blog/page/2/ and /blog/page/3/ at /blog/ tells Google those pages are copies of page 1. They aren't, so the articles linked only from deeper pages can lose their crawl path. Give each paginated page a self-referencing canonical.

9. hreflang pages canonicalized to one language

The French page's canonical points at the English page, so Google treats the French page as a duplicate and drops it. Each language version should canonical to itself and list the others with rel="alternate" hreflang.

10. Canonical added by JavaScript that differs from the HTML

If the raw HTML has one canonical and JavaScript rewrites it to another, you're sending two signals. Output the correct canonical server-side.

11. Using the URL removal tool to pick a canonical

Removing a duplicate with Search Console's Removals tool hides it temporarily, but Google says not to use the URL removal tool for canonicalization. It can hide every version of the URL from search, including the one you want. Use a redirect or a canonical tag instead.

How to see what Google chose

1. URL Inspection in Search Console shows the User-declared canonical next to the Google-selected canonical. If they differ, your signals are disagreeing.
2. The Page indexing report groups affected URLs:
   • Alternate page with proper canonical tag: working as intended. No action needed.
   • Duplicate without user-selected canonical: add canonicals.
   • Duplicate, Google chose different canonical than user: find the conflicting signal using the list above.

When Google overrides your canonical, the cause is usually one of three things. The two pages aren't actually similar enough. Internal links and the sitemap favor the other URL. Or your preferred URL redirects, returns an error, or is weaker than the duplicate.

Canonicals and syndication

If another site republishes your article, ask them to add a canonical pointing to your original. Google has said a cross-domain canonical isn't the best fit for syndication, because the pages are often quite different once you include the partner's template. A noindex on the syndicated copy is the more reliable choice. At a minimum, ask for a clear link back to your original.

Frequently asked questions

What is the difference between rel=canonical and a canonical meta tag?

There is no separate canonical meta tag. "Canonical meta tag" is a loose name for the <link rel="canonical" href="..."> element. Google documents the link element, the HTTP Link header and sitemaps as ways to declare a canonical. A meta element named canonical is not one of them.

Does link rel canonical go in the head or the body?

The head. Google only accepts the rel="canonical" link element when it appears in the <head>. If a canonical shows up in the body, often because a script or a broken tag closed the head early, it is ignored. View the page source to confirm where it sits.

What is a canonical redirect?

It's a 301 (permanent) redirect that sends a duplicate URL to the preferred one, for example from http:// to https:// or from the non-www host to www. Unlike a canonical tag, the duplicate stops loading for visitors. Use it when nobody needs the old URL any more, and make sure the canonical tag on the target page names the same final URL.

Does a canonical tag pass link equity like a 301?

Google treats both as strong signals for consolidation, so signals from the duplicates generally move to the canonical. A redirect is the stronger and more predictable of the two. Use it when the old URL doesn't need to be reachable.

Should every page have a canonical tag?

Every indexable page should have a self-referencing one. It costs nothing and protects against parameter duplicates you don't control. Pages you've noindexed don't need one.

Can I canonical similar but not identical pages, like color variants?

Yes, if the main content is essentially the same and you only want one version ranking. If the variants have meaningfully different content that people search for, such as different models or specs, let each one be canonical to itself. Otherwise Google may ignore the hint anyway.

Why is Google ignoring my canonical tag?

Usually because other signals contradict it, the pages aren't true duplicates, or the target redirects, 404s or is noindexed. Compare the user-declared and Google-selected canonicals in URL Inspection, then check redirects, internal links, the sitemap and hreflang for that URL.

What to do next

Pick your ten most important pages and check each one for a single, absolute, self-referencing canonical that returns 200. The free meta tag checker shows the canonical a page declares alongside its title and robots tags. Then check that your sitemap and internal links use exactly the same URLs. Run a free audit to check the canonical on any page in seconds. For a site-wide view, the paid whole-site crawl flags canonical conflicts, canonicals that point to redirects or noindexed pages, and pages blocked from indexing.

Related reading

• How Long Does SEO Take to Work? A Realistic Timeline: How long does SEO take to work? Fixes can show in days; content and authority take months. A stage-by-stage timeline, what Google says and what speeds it up.

Sources

• Google Search Central: What is canonicalization
• Google Search Central: How to specify a canonical URL with rel="canonical" and other methods
• Google Search Central: Introduction to robots.txt
• Google Search Central: Build and submit a sitemap