Duplicate Without User-Selected Canonical: What It Means and Whether to Fix It

What a healthy page looks like

A page that names itself shows its own address on User-declared canonical. On Google-selected canonical, Search Console prints Inspected URL, meaning Google kept the address you inspected. Google's URL Inspection help, re-read on October 19, 2026, puts it this way: with no other versions, "the Google-selected canonical is the inspected URL." Below is theo.site's homepage, read through Google's URL Inspection API on October 18, 2026.

None of the 124 addresses in theo.site's sitemap was under this reason when Theo inspected each one that day; the full table is in the guide to Discovered - currently not indexed. Without Search Console at hand, the homepage check on the not-showing-up guide reads whether your homepage names its own address. It checks the homepage only.

Addresses that now forward

An address that forwards belongs under Page with redirect, not here. If one still shows in this list, inspect it and find its Last crawled date. If that date is before the day the forward went in, Google has not been back since, and the address moves to Page with redirect on its next visit. If the date is after, open the old address in a browser: it should land on the new page. If it opens the page at the old address instead, the forward is not on yet.

After a move, old addresses spread across reasons: forwards under Page with redirect, retired pages under Not found (404). This one holds copies the new site still opens at, plus the old site if it is still live. The post on redesigning a website without losing SEO walks through a move.

Google kept another spelling of your page

For a kept http:// address, Google's duplicate guide names what can change its usual choice: "Google prefers HTTPS pages over equivalent HTTP pages as canonical, except when there are issues or conflicting signals such as the following". In plain words, its list is: the https:// page's certificate is not valid, the page loads files other than pictures from http:// addresses, it forwards to or through http://, or its canonical tag names the http:// address. The guide also asks that the sitemap list the https:// version, not the http:// one.

For a kept www, /index.html or end-slash spelling, start at step 2.

1. Open the https:// address in a browser and check that the address bar shows a lock, or in Chrome a small icon of two sliders, and that the address stays on https://. If a warning or a Not secure label shows instead, the guide to a website that says Not secure covers each cause.
2. Change any links on your site that use the extra spelling, often an old menu link or button to an http:// or /index.html address.
3. Make sure your sitemap lists only main addresses, each starting with https:// and none ending in /index.html. Most builders keep it current on their own.
4. Have each page carry a canonical tag naming its own main https:// address. The copies then move to Alternate page with proper canonical tag, which needs nothing.
5. If your host or platform offers a permanent forward (a 301) from the extra spelling, turn it on. If it does not, the steps above give Google enough to go on.

Google kept a copy on another website

• Your old site is still live at its old host or builder address. Take it down, or forward each old address to the matching new page if the old host allows it.
• A supplier's or maker's text appears on many sites at once. Rewrite yours in your own words.
• A menu or brochure posted on a directory or partner's site can win over your page. Ask them to link to your page instead.

Pages meant to be different

When a listed page is not really a copy, Google's Page indexing help asks you to "ensure that the content differs substantially between the two pages." Naming a main copy is "a hint, not a rule", in the words of Google's canonicalization guide, so Google can still treat two near-identical pages as one. A page earns its own place with words, photos and answers only it has.

Town and service pages run into this most. Read two aloud: if a customer could not tell which town each serves without the name, they need more of their own. If each page already names itself, Google files them under Duplicate, Google chose different canonical than user instead, with the same fix.

When Theo's Growth Engine writes a town or service page, it compares the draft with the most similar page already on the site, and rewrites anything that reads like the same page with a name swapped before the page is done.

What not to do

• Blocking copies in robots.txt. A blocked page cannot be read, so Google never learns which page it copies. Google's guide says plainly: "Don't use the robots.txt file for canonicalization purposes."
• Adding noindex to steer Google's choice. Google's duplicate guide says: "We don't recommend using noindex to prevent selection of a canonical page within a single site, because it will completely block the page from Search. rel="canonical" link annotations are the preferred solution." Those annotations are the canonical tag this guide describes.
• Requesting indexing for a copy. That asks Google to show the address you want left out. Request it once, for the main page, after a change.

WordPress

WordPress itself writes a canonical tag on single posts and pages, and has since version 2.9. The WordPress developer reference shows that built-in tag is written only for one post or one page at a time. Category, tag and date lists, and page 2 onward of the blog, get none from WordPress itself.

Those archives repeat your posts' titles and first lines, so Google often files them here, and that is fine. A tag used on one post makes a page that nearly copies it, so fewer, broader tags keep the list shorter. Many themes and SEO add-ons add the tag to archives; view an archive's page source and search for canonical to see.

Sites built by hand or in a desktop program

Sites made in Dreamweaver, Mobirise or a similar program carry no canonical tag unless someone adds one. With no tag, any second address for a page lands here. The fix is one line in the head of each page, naming the page's main address, the https:// one without /index.html. A copy of a page names the main page instead. On a services page it reads <link rel="canonical" href="https://yourbusiness.com/services.html">. On the home page, name the plain address, https://yourbusiness.com/, not /index.html.

In Mobirise, open the Pages tab, click the gear button on a page, and paste that page's line into the box named Inside <head> code. Mobirise's help on Page Settings, re-read on October 19, 2026, says that box comes with its Code Editor extension. Do it page by page, then publish again. The Analytics box in Site Settings will not do: its help on Site Settings says it adds the same code to every page.

In Dreamweaver, open each page, switch to Code view, and paste its line anywhere between <head> and </head>. Then upload the changed pages to your host as usual.

One old tip says to set a preferred www version in Search Console; Google retired that setting in June 2019.

A note for whoever runs your site

Shopify, Wix, Squarespace, GoDaddy and Mobirise

Shopify, Wix and Squarespace write the tag for you, so this reason is rare on them. GoDaddy's help does not say; view a page's source and search for canonical. Mobirise's help lists no canonical option in its settings, so the line goes in page by page, as above. When the reason appears on a builder, check whether the kept address is the builder's own free address or a copy on another website; the address checker reads both.