Broken Link Checker

This broken link checker crawls one site and lists dead links: each URL that returned a 4xx or 5xx status, with the page that linked to it.

Crawls up to 200 URLs on that host, then checks linked URLs until the same limit. Private and local addresses are not requested.

How to check a site for broken links

  1. Enter your homepage. Use the public https address. We crawl that host, including a www redirect from the address you typed, and we do not crawl other sites.
  2. Wait for the crawl. We request up to 200 URLs. robots.txt disallow rules on your host are skipped. Private and local addresses are refused.
  3. Read the 4xx and 5xx rows. Each row is a link that returned an error, plus the page where we found it. A timeout stays off the list because we never got a status code.

What gets listed

A broken link here is a response code from 400 through 599. A 301 or 302 that lands on a 200 is not broken. The checker follows up to four hops. The source column is a page on your site whose HTML contained the link. The starting URL is listed on its own when that address itself returns an error.

The crawl reads the HTML the server sends. It does not run scripts, fill in forms, or sign in. A menu that only appears after JavaScript loads will not be in the report. Links inside script and style blocks are ignored.

What to fix first

Start with 404 and 410 links on your own host. Point them at the page that replaced them, or remove the link. A 5xx error is a server problem, not a typo in the href. Check one address hop by hop when the chain itself looks wrong.

Build the URL list with the XML sitemap generator. See every anchor on a single page with the site link analyzer. Follow one redirect chain with the redirect checker.

Questions

What is a broken link checker?

It requests the links on your pages and lists the ones that return a 4xx or 5xx status. 404 means not found, 410 means gone, and 5xx means the server failed. The row includes the page that linked there, so you can edit that page.

Is a dead link checker different?

On this page, dead link and broken link mean the same thing: the target responded with 4xx or 5xx. A link we could not reach, such as a timeout, is counted as not checked. It is not labeled dead, because we did not get an error status.

Do you crawl the whole site and every external site?

No. The page crawl stays on your host and stops at 200 requests, including robots.txt and redirects. External links are checked once, if there is room left, and we do not follow links from those other sites. Pages added only by JavaScript are not seen.

Why would a working page show 403?

Some servers refuse crawlers and still show the page to a browser. 401 and 403 are listed because they are 4xx responses. Open the URL yourself before you delete it.

Do you respect robots.txt?

Yes, on the host you asked us to crawl. A disallowed URL is not requested, so it is not called broken. If robots.txt returns a server error, the crawl stops. We do not download a separate robots.txt for every external site a link points at.

Will this request private or internal addresses?

No. Localhost, .local, .internal, and private, loopback, and link-local IP addresses are refused before any request. Only http port 80 and https port 443 are used.

More free tools

See all tools