Canonical Tags and Duplicate Content: How Google Picks a URL
On this page
Duplicate content is the same, or nearly the same, content available at more than one URL. When Google finds duplicates, it groups them and picks one canonical URL to show in search results. You can tell Google which version you prefer with redirects, a rel=canonical tag or your sitemap, but Google treats your choice as a strong hint, not a rule. This explainer is part of our technical SEO guide.
Key takeaways
- Duplicate content is the same content at more than one URL; some is normal.
- Google groups duplicates and picks one canonical URL to show in search.
- Redirects and rel=canonical are strong signals; sitemap inclusion is a weak one.
- Your canonical is a hint, not a rule; Google can choose a different URL.
- URL Inspection shows both your declared canonical and the one Google selected.
What duplicate content is, and why it's normal
Most duplicate content isn't copied on purpose. It comes from the way websites work. Google lists the usual causes: HTTP and HTTPS versions of the same page, separate mobile and desktop URLs, regional versions in the same language, sorting and filtering options that create new URLs, and test or demo versions left open to crawlers.
Small differences in a URL count too. A trailing slash, capital letters, a tracking parameter or a session ID can each turn one page into several addresses with the same content.
Google says some duplicate content on a site is normal and isn't against its spam policies. The problems are practical. Links and other signals are split across several URLs instead of building up on one. Crawlers spend time on copies. And your reports show the same page under different addresses, which makes performance harder to read.
How Google chooses the canonical URL
When Google indexes a page, it works out the page's main content. If it finds other pages whose main content is the same or very similar, it clusters them. It then picks the one it judges most complete and useful for searchers and marks it as the canonical.
The canonical gets crawled most often and is the main version Google uses to evaluate the content. The duplicates are crawled less, and search results usually point to the canonical. Google may still show a duplicate when it suits the searcher better, such as the mobile version of a page on a phone.
Google weighs several factors when it chooses: whether a page uses HTTPS, redirects, whether the URL is in a sitemap, and rel=canonical annotations. It prefers HTTPS pages by default, unless something such as a bad certificate gets in the way.
The signals you can send, strongest first
You don't have to declare a canonical. If you don't, Google picks one itself. But stating your preference helps when you care which URL people see. Google ranks the methods by how strongly they influence its choice, and they work best when they agree.
| Signal | Strength | Best used for |
|---|---|---|
| Permanent redirect | Strong | Duplicates that should no longer exist |
| rel=canonical link or header | Strong | Variants that must stay reachable |
| Sitemap inclusion | Weak | Stating preferences across a large site |
| Internal links to one version | Supporting | Every page, all the time |
The rel=canonical link element sits in the page's head and names the preferred URL. Use the full, absolute address:
<link rel="canonical" href="https://www.example.com/services/roof-repair">Put the same element on the canonical page itself, pointing to its own URL. Google recommends this self-referencing canonical. For files such as PDFs, which have no head section, the same instruction can be sent as an HTTP header. List only canonical URLs in your XML sitemap, and point your internal links at the canonical version too.
Canonical mistakes to avoid
Google's guidance names several mistakes that send mixed or harmful signals.
- Using robots.txt to hide duplicates. Google can still index a blocked URL without seeing its content, and it can't see the canonical tag on a page it can't crawl.
- Using noindex to pick a canonical. Noindex removes the page from search completely. Google recommends rel=canonical instead.
- Sending conflicting signals. Don't list one URL in your sitemap and name a different one in the canonical tag for the same page.
- Pointing to a URL fragment. Google generally doesn't support the part of a URL after a # sign as a canonical.
- Letting JavaScript change the canonical. Set it in the HTML source and keep scripts from rewriting it.
- Using the URL removal tool. It hides every version of the URL from search, not just the duplicate.
How to check which URL Google chose
The URL Inspection tool in Search Console shows two fields under Page indexing: the user-declared canonical, which is what your page asks for, and the Google-selected canonical, which is what Google chose. If they match, your signals are working. Google notes that only the indexed result shows its choice; a live test can't predict it.
The Page indexing report groups duplicates into statuses. "Alternate page with proper canonical tag" means the page points to its canonical correctly, so there is nothing to fix. "Duplicate without user-selected canonical" means Google picked a canonical because you didn't. "Duplicate, Google chose different canonical than user" means Google overruled you. Our guide to indexing statuses covers each one.
When Google picks a different URL than you, first ask whether its choice actually serves searchers better. If not, make your signals agree, and if the pages are genuinely different, make that difference clear in the content. Google says it may keep pages clustered for up to two weeks after a fix. If you'd rather have duplicate URLs untangled for you, our in-house SEO team handles site structure and metadata.
Common questions
Will duplicate content get my site penalized?
Not by itself. Google says some duplicate content on a site is normal and isn't a violation of its spam policies. Google simply groups the duplicates and shows one version. The real costs are practical: links and other signals get split across URLs, crawling is wasted on copies, and your reports are harder to read. Copying other sites' content to manipulate rankings is a different matter.
Should every page have a self-referencing canonical tag?
Google recommends it. Adding a rel=canonical link that points to the page's own URL makes your preference explicit, so a version reached with tracking parameters or a different letter case still points back to the main URL. Use the full, absolute URL, keep only one canonical per page, and make sure JavaScript doesn't change it after the page loads.
Is a canonical tag the same as a redirect?
No. A redirect sends people and crawlers to another URL, so the old one is no longer visited. A canonical tag leaves both pages reachable and asks Google to treat one as the main version. Use a redirect when a duplicate page should disappear, such as an old URL after a move. Use a canonical when the variant must stay available, such as a filtered or sorted view.