Paginated pages canonicalised to page one: how deeper posts and products get lost
Page two of a category is not a duplicate of page one: it lists different posts or products. A canonical from /category/mugs/page/2/ to /category/mugs/ asks Google to disregard page two, and with it the links to everything listed there. Google's guidance is to give every page in a paginated series its own self-referencing canonical.
The mistake is usually well-intentioned. Someone noticed paginated URLs in a duplicate-title report and "consolidated" them, or a theme or plugin setting did it wholesale.
What it looks like
Before
<!-- on https://example.com/category/mugs/page/2/ -->
<link rel="canonical" href="https://example.com/category/mugs/">
After
<!-- on https://example.com/category/mugs/page/2/ -->
<link rel="canonical" href="https://example.com/category/mugs/page/2/">
What goes wrong
A canonical says two URLs hold the same content. When Google accepts that for page two, it has little reason to keep crawling page two or to count the links on it. For items that are linked only from deeper listing pages, such as older blog posts and products far down a category, those links were their main route to being discovered and re-crawled. Sites with this setup often see old content drift into "Discovered - currently not indexed" or become orphan pages in practice.
Google may also notice that the pages are plainly different and ignore the tag, in which case the canonical is doing nothing. Neither outcome helps.
Check your archives
Request page two of a few listing types and read the tag:
for u in /blog/page/2/ /category/news/page/2/ /shop/page/2/ /author/editor/page/2/; do
echo "$u -> $(curl -s "https://example.com$u" | grep -io '<link[^>]*rel=.canonical.[^>]*>')"
done
Each line should show a canonical that ends in /page/2/. Check multi-page posts too, the kind split with the Page Break block: /long-guide/2/ should canonicalise to /long-guide/2/.
What WordPress and Hydrogen SEO do by default:
- WordPress core prints a canonical only on single posts and pages, and for a multi-page post it includes the page number. It prints no canonical on archives.
- Hydrogen SEO prints a self-referencing canonical on archives as well, and appends the
/page/N/segment on paged views, so each page in a series names itself.
If you see page one as the canonical on deeper pages, look for a theme function, a custom wp_head snippet, a pagination or "load more" plugin, or a per-term canonical override that is being applied to every page of the term's archive.
The correct setup for a series
- Self-referencing canonical on every page. Page N points to page N.
- Crawlable links between pages. Real
<a href>links to the next and previous pages, or numbered links. A "load more" button with no underlying URL hides everything after the first batch from crawlers. - Leave the pages indexable.
noindexon paginated pages has a similar long-term effect to the wrong canonical: pages that stay noindexed are crawled less, and their links count for less. - Do not block
/page/in robots.txt, for the same reason.
rel="next" and rel="prev" link tags are harmless, but Google announced in 2019 that it no longer uses them, so they are not a substitute for any of the above.
Duplicate titles are a separate, smaller matter
Paginated pages often share one title and description, and that is what audit tools flag. Solve that directly instead of with canonicals. Where your theme or SEO setup allows it, add the page number to the title of pages two onward. Google rarely shows deeper listing pages for a query anyway, so identical titles here cause little harm, but distinct ones keep reports clean.
When a "view all" page exists
There is one case where a canonical away from paginated pages is right: you have a single view-all page containing every item in the series, it loads acceptably fast, and you would prefer it to be the result shown. Then the paginated pages can canonicalise to the view-all URL, because its content really does include theirs. This does not apply to canonicalising to page one, which contains only the first batch.
Reduce reliance on pagination
Fixing the tag restores the baseline. To make deep items easier to reach, shorten the path: more items per page, subcategories that split long archives, links from related posts, and a hub page for evergreen content. Hydrogen SEO's internal link counter shows how many links each post receives and has an orphaned filter for posts with none, which is a quick way to see what your archives are failing to surface.
Confirm each paged URL names itself
Re-run the loop after clearing caches and confirm each paged URL names itself. In URL Inspection, test /page/2/ of an important archive: the user-declared canonical should match the inspected URL. Over the following weeks, deeper posts should show more recent crawl dates when inspected.
Common questions
Is page two duplicate content?
No. It has the same template and often the same title, but its main content, the list of items, is different.
Should I noindex paginated pages instead?
We would not. It is unnecessary, and over time it reduces how often those pages are crawled, which weakens the links to the items they list.
Does Google still use rel=next and rel=prev?
No. Google said in 2019 that it had stopped using them as an indexing signal. Other search engines and browsers may still read them.
Will paginated pages compete with page one in results?
Rarely. Google generally shows the first page for relevant queries because it has the strongest internal links.