Short answer
The canonical URL is the address Google picks as most representative from a set of similar pages and keeps in its index. You state your preference with redirects, rel=canonical and a sitemap; these are hints, not rules. You can see Google’s choice in Search Console URL Inspection.
The owner of a small boutique notices that visitors from a campaign email reach the green dress page at an address with ?utm_source=newsletter added. The same dress also opens with and without www, and again through a colour filter. Which one shows up on Google? The canonical tag, meaning the declaration of a canonical URL, is where your site gives its answer. This article explains how Google chooses the canonical address, which signals fix duplicate content, and how to see the result in Search Console.
What a canonical tag and duplicate content are
When the same or very similar content opens at more than one address, it is called duplicate content. According to Google’s canonicalisation documentation, Google groups these pages into a cluster and marks the one it judges most complete and useful for searchers as the canonical page. The canonical page is crawled most regularly; duplicates are crawled less often.
The documentation is clear on two points. First, some duplicate content on a site is normal and does not violate Google’s spam policies. Second, stating a canonical preference is a hint, not a rule: Google may choose a different page than you did.
The rel="canonical" tag is a link element placed in the page’s <head> that points to your preferred address. It is not the only signal, though. Below we look at three of them and how strong each one is.
How Google chooses the canonical address
Google’s canonical URL documentation ranks the methods by influence: a redirect is a strong signal, a rel="canonical" link is a strong signal, and including a URL in a sitemap is a weak signal. The methods stack, so using two together raises the chance that your preferred URL is chosen. None of them is required; if you use none, Google picks a version itself.
Signals from the site’s setup also count. The same document says Google prefers an HTTPS page over the equivalent HTTP page, unless there is a problem such as an invalid certificate.
According to Google’s documentation, your site will probably do fine without a canonical preference, because Google picks the best version itself. The value of stating one grows with the number of products and parameters. Google’s reasons include choosing which address appears in results, consolidating links on one URL, simplifying metrics and not spending crawl time on duplicates.
It matters that the signals agree and point to the same address. The document advises against giving different URLs for the same page through different methods, for example one address in the sitemap and another in rel="canonical".
Typical causes of duplicate content
First find out where the copies come from. Google’s documents list region and device variants, HTTP and HTTPS versions, site functions such as sorting and filtering, and demo sites accidentally left open to crawlers. The table below maps these causes to what they look like on a typical shop or company site, with the suggested fix.
| Cause | Example | Suggested fix |
|---|---|---|
| Protocol and host | http:// and https://, with and without www |
Pick one version and 301-redirect the others |
| Trailing slash | /dress and /dress/ |
Our suggestion: pick one form and 301-redirect the other |
| Tracking parameters | ?utm_source=newsletter, ?gclid=... |
rel="canonical" pointing to the clean address |
| Sorting and filters | ?order=price, ?colour=green |
Keep out of the index with noindex or robots.txt |
| Pagination | ?page=2, ?page=3 |
Each page carries its own canonical address |
| Device and region variants | Separate mobile address, two country pages in one language | rel="canonical" and, where needed, hreflang |
For the sorting and filter row, Google’s pagination documentation suggests the noindex robots meta tag, or discouraging crawling with robots.txt, so that variations of the same list are not indexed. The same document says not to make the first page the canonical for the whole sequence, and to give every page its own canonical URL. Google no longer uses rel="next" and rel="prev".
Example: one product, five addresses
Take the green dress page of the shop above (the addresses are examples). Five addresses serve the same content, and each needs one decision:
http://example.com/green-dressis 301-redirected to the HTTPS, www version.https://example.com/green-dressis 301-redirected to the www version.https://www.example.com/green-dressgoes by 301 to the form with a trailing slash, because the shop chose that form.https://www.example.com/green-dress/?utm_source=newsletterkeeps working for the campaign link, and the page carries arel="canonical"pointing to the preferred address.https://www.example.com/dresses/?colour=greenis a filtered list;noindexor a crawl restriction is applied to it.
In the end every signal sent to Google (redirect, canonical, sitemap, internal link) points to the same single address. That agreement does more than the strength of any one signal.
Step by step: fix duplicate content
1. List every address of a page
Pick your five busiest pages. For each, try the http/https, www, trailing slash and known parameter variants by hand in the browser. Write down which addresses return the same content. Also check how your site treats the trailing slash, upper and lower case, and parameter order; these are easy to miss. Finding even one extra address explains why Search Console shows two addresses for the page.
2. Choose one preferred address per page
Choose the readable, parameter-free, HTTPS form that is consistent with the rest of the site. This address will be used unchanged in the sitemap, internal links and tags.
3. Redirect the addresses that should not exist
According to Google’s documentation, permanent redirects are the right method when you want to get rid of duplicate pages, and server-side (HTTP) redirects take effect fastest. Finish the moves from http to https and from non-www to www (or the reverse) in a single 301 step.
4. Add rel=canonical to copies that must stay reachable
For variants that have to stay open, such as addresses with campaign parameters, put a link in the <head> that points to your preferred address:
<link rel="canonical" href="https://www.example.com/green-dress/" />
Google’s recommendations are these: write the full (absolute) address, place the element in the <head>, and add a self-referencing canonical to the canonical page itself. Choose either the HTML element or the HTTP header; using both is error-prone. The header helps with non-HTML files such as PDFs. If you add the tag with JavaScript, make sure it is injected correctly.
5. Align the sitemap and internal links
Put only canonical addresses in the sitemap; according to Google, all pages listed in a sitemap are suggested as canonicals. Link internally to the canonical address rather than a duplicate, because the document says this makes your preference clearer. For sitemap and robots.txt preparation see our robots.txt and sitemap guide, and for link structure our internal linking article.
6. Check Google’s choice in URL Inspection
Paste the page address into the search box at the top of Search Console. According to the URL Inspection tool documentation, you can determine the canonical only in the indexed data; the live test cannot predict whether the tested version will be considered canonical. In the Page indexing section two fields sit side by side: “User-declared canonical” and “Google-selected canonical”.
- If they match, your preference was accepted.
- If the user-declared field says “None” and the page has no alternate versions, that is fine.
- If they differ, go to the next step.
7. If Google chose another address, look for the reason
Google’s troubleshooting documentation first asks whether Google’s choice makes more sense for your users than yours. Then look at technical causes: a CMS or plugin may be pointing at an unwanted address. Check also whether the clustered pages are really different enough. After a fix, Google may hold pages in the duplicate cluster for up to two weeks. For important addresses, request indexing from URL Inspection, but because there is a quota, keep it for your most important pages.
Common mistakes
- Using robots.txt for canonicalisation. Google advises against it; disallowed URLs may still be indexed without their content.
- Conflicting signals. One address in the sitemap, another in the canonical tag, or a canonical and a 301 pointing in different directions.
- Using noindex to state the canonical version. Google does not recommend it; the page disappears from search entirely. Within one site the preferred solution is
rel="canonical". - Using relative addresses. Supported, but they cause trouble if a test site is crawled by accident; write absolute addresses.
- Pointing every paginated page at the first page. Each page should carry its own canonical address.
- An HTTPS page that points to HTTP. Google’s document lists this among the conflicting signals that are exceptions to its HTTPS preference.
- Pointing the canonical of a multilingual site at one language. For language versions, see the rules in our hreflang article.
- Not checking the tag a plugin generates. The troubleshooting document says some CMSs and plugins can point to the wrong address.
This check is one part of a general technical audit; for the other items see our technical SEO checklist. For catalogues with many filters and parameters, our e-commerce SEO article is the next stop. If you would rather have us audit and fix it, we do that as part of our SEO service.
Common questions
Does a canonical tag replace a redirect?
No. A redirect takes the visitor and Google to another address; a canonical states a preference while both addresses stay. According to Google, a redirect is for when you want to get rid of the duplicate. If the copy has to stay reachable, a canonical fits.
Will I be penalised for duplicate content?
According to Google’s documentation, some duplicate content on a site is normal and does not violate its spam policies. The real problems are that users cannot tell which page is the right one and that performance is harder to track.
Should every page carry a self-referencing canonical?
Google recommends adding a self-referencing rel="canonical" to the canonical page itself. Parameter copies then point to the same address as the canonical page.
One thing you can do today: paste the address of your busiest page into Search Console URL Inspection and compare the two canonical fields. If they differ, try the http, www and parameter variants of that page by hand and note which one comes from where.
Sources
- Google Search Central: How to specify a canonical URL with rel="canonical" and other methods
- Google Search Central: What is canonicalization
- Google Search Central: Fix canonicalization issues
- Google Search Central: Pagination, incremental page loading, and their impact on Google Search
- Search Console Help: URL Inspection tool
This article is general information, not legal, financial or technical advice. Platforms, their rules and the law change over time, so some details may have changed since 15 November 2026 or contain small errors. Before you make a decision for your business, check the current official sources or write to us.