Quick answer: Pagination SEO is about helping Google understand that page 1, page 2 and page 3 of a list belong to one series. Add rel=”next” and rel=”prev” link elements in the head of each paginated page, give every page a self-referencing canonical (not a canonical pointing back to page 1), keep the pages crawlable with plain HTML links, and avoid noindexing the series by default. If you offer a “view all” page that loads quickly, you can canonicalise the series to it instead.
Most sites have pagination somewhere. Blog archives, category pages in a shop, search results, forum threads and long articles split across several pages all use it. When it is set up badly, you get thin duplicate pages in the index, products that never get crawled, and ranking signals spread across URLs that nobody wants to land on. When it is set up well, Google crawls deep into your catalogue and sends visitors to the most useful page in the set.
This guide walks through how pagination works for search, how to implement rel=next and rel=prev, and the common mistakes we see in audits.
Why pagination causes SEO problems
A paginated series creates several URLs that look very similar. Page 2 of a category often has the same title, the same meta description and the same intro text as page 1, with only the list of items changing. To a crawler that can look like near-duplicate content.
There are three separate risks:
- Duplicate signals. Titles and descriptions repeat across the series, so Google has to guess which page matters.
- Crawl depth. Products or posts that only appear on page 9 of a category are many clicks from the homepage. If the series is blocked or broken, those items may never be found.
- Diluted relevance. Links and engagement get split across page 1, page 2 and so on instead of building up one strong URL.
Pagination markup does not make these risks disappear, but it gives Google a clear map of the series.
How rel=next and rel=prev work
Google introduced support for rel=”next” and rel=”prev” in 2011. The idea is simple: each page in a series tells search engines which page comes before it and which comes after it. Google can then treat the series as a logical sequence, consolidate signals, and usually send searchers to the first page.
The markup goes in the head of the document:
Page 1 (example.com/blog/) has only a next link:
<link rel="next" href="https://example.com/blog/page/2/">
Page 2 has both:
<link rel="prev" href="https://example.com/blog/">
<link rel="next" href="https://example.com/blog/page/3/">
The last page has only a prev link:
<link rel="prev" href="https://example.com/blog/page/9/">
A few details matter:
- Use absolute URLs to avoid confusion with relative paths.
- Point page 2’s prev link at the clean page 1 URL, not at /page/1/, if your site redirects /page/1/ to the root.
- Keep URL parameters consistent. If page 2 is ?page=2&sort=price, the next link should be ?page=3&sort=price, not ?page=3 on its own.
Google has said these links are a hint, not a directive. They help, but they work best alongside sensible canonicals and internal linking.
Canonical tags on paginated pages
This is where most implementations go wrong. A common mistake is to set the canonical on every paginated page to page 1. That tells Google that page 2, page 3 and so on are duplicates of page 1, which they are not. The content on page 5 is different, and if Google follows the canonical it may stop indexing the items that only appear there.
The recommended approach:
- Each page canonicalises to itself. Page 2’s canonical is page 2’s own URL.
- rel=next and rel=prev connect the series.
- Parameter variations canonicalise to the clean version. If page 2 can be reached with tracking parameters or session IDs, those versions should point their canonical to the clean page 2 URL.
You can combine rel=canonical and rel=next/prev on the same page. Google has confirmed they are independent signals.
The “view all” option
If you can offer a single page that shows the full list and it loads quickly, Google has said it may prefer to show that page in results. In that case, add a canonical from each paginated page to the view-all page and skip rel=next and rel=prev.
This works for a 40-item article split into pages. It does not work for a category with 2,000 products, because a page that big will be slow and painful on mobile. With mobile-first indexing now rolling out, a heavy view-all page is a poor choice for large catalogues.
Should you noindex paginated pages?
Some SEO plugins and older guides suggest adding noindex, follow to page 2 onwards. The thinking is that it keeps thin pages out of the index while still letting Google follow the links.
We rarely recommend it as a default. If pages are set to noindex for a long time, Google may eventually crawl them less and pass less value through their links, which defeats the point of keeping them followable. With proper rel=next/prev and self-referencing canonicals, there is usually no need to hide the series.
Noindex can make sense for internal search result pages, which you generally do not want in the index at all, or for filtered combinations that create endless near-identical URLs. That is a faceted navigation problem more than a pagination problem.
Make the series crawlable
Markup in the head is only part of it. Google also needs real links it can follow.
- Use plain anchor links for page numbers and next/previous buttons. Pagination that only works through JavaScript click handlers or form posts can stop crawlers at page 1.
- Infinite scroll needs a fallback. If your blog or shop loads more items as the user scrolls, make sure each chunk also exists at a paginated URL, and update the URL with the History API as the user scrolls. Google has published guidance on making infinite scroll search-friendly this way.
- Link to more than just “next”. Numbered links (1, 2, 3 … 9) let crawlers reach deep pages in fewer hops than a single next button.
- Do not block pagination in robots.txt. Disallowing /page/ is a quick way to hide most of your archive.
For more on keeping crawl paths open, our technical SEO checklist covers related issues like orphan pages and redirect chains.
Titles, descriptions and on-page content
Even with good markup, small changes reduce the duplicate feel of a series:
- Add the page number to the title: “Running Shoes, Page 2 | Store Name”.
- Either vary the meta description by page or leave it empty on deeper pages so Google writes its own snippet.
- Show the category intro text only on page 1. Repeating 300 words of copy on every page of a series adds duplication without helping anyone.
For shops, category pages carry a lot of weight, so this ties in with broader category page SEO work like headings, filters and product grid layout.
Pagination in WordPress
WordPress paginates archives out of the box using /page/2/ style URLs. Core does not add rel=next and rel=prev to archives on its own, but Yoast SEO and similar plugins output them for blog, category, tag and author archives. Check your theme too: some themes add their own pagination links, and you can end up with duplicates.
Things to check on a WordPress site:
- View source on page 2 of your blog and confirm there is one prev, one next, and a canonical pointing to page 2 itself.
- If a post is split with the nextpage tag, check that the plugin handles paginated posts (/post-name/2/) as well as archives.
- Make sure the “Blog pages show at most” setting in Reading is sensible. Showing 10 posts per page with 900 posts creates 90 archive pages. Showing 20 halves that.
Pagination for ecommerce categories
Shops are where pagination matters most, because products deep in a category are the ones most likely to be missed.
- Combine pagination with good internal linking: subcategories, featured product blocks and related product links all reduce reliance on page 14 of a list.
- Keep sort and filter parameters out of the index unless a filtered view targets a real search term. Use canonicals pointing at the unfiltered paginated URL.
- Watch out for pagination that resets when filters are applied. If page 3 with a colour filter becomes page 3 of the full list, you have broken URLs.
- Check that out-of-stock or deleted items do not leave empty final pages that return 200 with no products.
How to audit your pagination
A quick audit takes less than an hour on most sites:
- Crawl the site with a tool such as Screaming Frog and filter for URLs containing /page/ or ?page=.
- Check the rel=next/prev and canonical columns for each paginated URL.
- Look for canonicals pointing to page 1 or to unrelated pages.
- Confirm that the final page in each series has no next link and that page 1 has no prev link.
- Search Google with site: plus a deep category URL to see what is indexed.
- In Search Console, check the HTML Improvements report for duplicate titles and descriptions caused by pagination.
FAQ
Does Google still use rel=next and rel=prev?
Yes. Google’s documentation recommends rel=next and rel=prev for paginated content, alongside self-referencing canonicals. Treat them as a hint that helps Google group the series.
Should page 2 have a canonical to page 1?
No. Page 2 shows different items, so its canonical should point to itself. Point it to page 1 only if page 1 really contains everything (for example, a view-all page).
Do Bing and other search engines read these tags?
Each engine handles pagination in its own way, but crawlable links, self-referencing canonicals and clear next/prev markup are a sensible setup for all of them.
Is infinite scroll bad for SEO?
Not if each loaded chunk also exists at its own crawlable, paginated URL. Without that fallback, content beyond the first load may not be indexed.
Getting pagination right
Pagination is one of those details that rarely gets attention until a crawl reveals hundreds of near-duplicate pages or a category where half the products are not indexed. Setting up rel=next and rel=prev, sensible canonicals and crawlable links is a small job that protects a lot of content.
If you want a second pair of eyes on your archives and categories, our search visibility service includes a crawl review that covers pagination, canonicals and crawl depth.
