Find the broken links on any page.

We fetch the page and collect every <a href> (and every image, if you tick the box), resolve them to full URLs, drop duplicates and skip mailto:, tel: and javascript: links. Each link gets a lightweight HEAD request; servers that refuse HEAD get a GET instead, because plenty of them answer HEAD wrongly. We check eight at a time with an 8-second limit per link.

1. 1Broken

   4xx and 5xx responses, redirect loops and too many redirects.

2. 2Can't connect

   Domains that don't resolve, refused connections and certificate errors.

3. 3Redirects

   Links that work but go through a redirect, with the final URL shown.

4. 4May block bots

   403, 429 and 999 answers that usually mean bot protection, not a dead page.

5. 5Timed out

   No answer within 8 seconds. Often a slow server rather than a dead one.

6. 6Context

   Anchor text, where on the page it sits (nav, main, footer), internal or external, nofollow.

Internal links first. A broken link to your own page is either a typo or a page that moved. Point it at the right URL, and add a 301 from the old one so links from elsewhere still work. Check the redirect with our redirect checker.

Navigation and footer links count most. They appear on every page, so one broken footer link is hundreds of broken links across the site. The "Where" column shows you which ones those are.

External links: replace dead ones with a live source, or link to an archived copy on the Wayback Machine. If a source is gone for good, remove the link rather than leaving readers at a 404.

AI assistants favour pages that look maintained and trustworthy, and a page whose sources 404 looks neither. Broken internal links also cut off the paths crawlers use to discover your other pages, including AI crawlers that never see your sitemap. Keep the links on your most-cited pages working, and make sure the URLs in your sitemap resolve too.