How to find broken links on your website

By · Updated

A broken link sends a visitor to an error page and wastes the trust that got them to click. Sites collect them quietly as other pages move or disappear. You can find them on any page in about a minute. This guide covers how to check, what the results mean, and what to do with each kind, using a real scan of one of our own pages as the example.

How to run the check

When we ran it on our own About page, it checked 82 links: 82 OK, 0 broken, 0 redirects and 0 blocked. This tool has to use our server to fetch the links, because a browser page is not allowed to test other sites directly. It sends the address you enter and stores nothing: our server fetches the page and each link only to read its status, and does not save the address or the results.

  • Open the Broken Link Checker and enter the address of one page in Page address (URL).
  • Click Check links. It tests every link and image on that page, up to 300, and shows progress as it goes. Click Stop if you want to end early.
  • Read the Problems tab first. The Redirects, OK and All tabs show the rest.
  • Download the CSV if you want a list to work through. It includes each URL, its status code, the result, the final address after any redirects, the link text, whether it is a link or an image, and whether it is internal or external.
  • Repeat for your other important pages.

What the results mean

Each link ends up in one of the groups in the table below. Under every result you can also see whether it is a link or an image, its link text, whether it points to your own site (Internal) or elsewhere (External), and, for redirects, where it ends up.

Blocked is not the same as broken. Plenty of large sites turn away automated requests while working perfectly for people, so check those by hand. A link that redirects and then lands on a missing page counts as broken, because what matters is where the visitor ends up.

ResultMeaningWhat to do
OK The link works Nothing
Broken A missing page (such as 404 or 410), a server error (500 and up) or a domain that no longer exists Update the link or remove it
Redirect It still works, through another address Point it at the final address when convenient
Blocked The site refused an automated check, with 401, 403 or 429 Open it in a browser before changing anything
No answer No response in time, or too many redirects Try the link yourself; the site may be slow or down for now

A worked example: triaging one page

Suppose a blog post has 40 links and the checker reports 34 OK, 3 redirects, 2 broken and 1 blocked. Here is a sensible order to handle them, most urgent first.

  • The 2 broken links. Open each one. If the page moved, search the site for its new address and update your link. If it is gone for good, find another source or reword the sentence.
  • The 1 blocked link. Open it in your browser. If it loads normally, the site simply refused our automated check and there is nothing to fix.
  • The 3 redirects. Look at where each one ends up. If http:// now goes to https://, or an old address goes to the same article at a new address, update the link to the final address when you next edit the post. If a redirect lands on a site's home page instead of the article you linked, treat it as broken: the content you cited is probably gone.
  • Run the check again to confirm the Problems tab is empty.

Which pages to check first

The tool checks one page at a time, so start where a broken link costs the most: your home page, main menu and footer (which appear everywhere), your most visited articles, and any page that earns money or leads.

Old articles collect the most dead links, because they often point to sources that have since moved. Check them once or twice a year. A link in your menu or footer appears on every page, so fixing it once on the template fixes it everywhere.

How to fix what you find

Broken outgoing links are mainly a problem for visitors. Google has said broken links are a normal part of the web, so do not panic over a few. Treat this as upkeep. To see how strong the sites linking to you are, try the Bulk Domain Authority Checker.

  • If the page moved, update the link to its new address.
  • If it is gone, find a replacement source or remove the link and adjust the sentence.
  • If the broken page is on your own site, add a redirect to the closest live page so existing links keep working.
  • Never delete a page that other sites link to without a redirect: that throws those links away.
  • For a broken image, upload the image again or point to a working copy, and keep its alt text.

What the checker does not see

Knowing the limits helps you read a clean result correctly. The checker reads the page's HTML as the server sends it, the same way a search engine first sees it.

  • Links added by JavaScript after the page loads are not in that HTML, so they are not checked.
  • Only ordinary links and images are tested. Email (mailto:), phone (tel:) and javascript: links are skipped, as are links to a #section on the same page.
  • The part of an address after # is ignored, so a link to a missing section of a working page shows as OK.
  • A page that says "not found" but returns a normal success code (a so-called soft 404) shows as OK, because the server reported success.
  • The same address linked twice on a page is checked once.
  • Only public websites can be checked. Addresses on your own computer or a private office network are refused.
  • On a page with more than 300 links, the first 300 are checked and a note tells you the rest were skipped.

If the page itself cannot be checked

Sometimes the tool cannot read the page you entered, so it has no links to test. The message tells you why. If you run many checks in a short time you may be asked to wait a minute; the limits keep our server from hammering other sites.

MessageLikely causeWhat to do
That website could not be found A typo in the address, or the domain has expired Check the spelling
The page answered with error 404 (or another code) The page you entered is itself missing or broken Open it in a browser and check the address
The website did not respond The site is down, slow or blocking automated checks Try again later
That address is not a web page (HTML) You entered a PDF, image or other file Enter the page that links to it instead
This address can't be checked A local, private or unusual address or port Use the public address of the page

A simple routine

Most sites only need a light, regular check rather than a big clean-up. Once a quarter, run the Broken Link Checker on your home page and your five most visited pages, fix anything on the Problems tab, and keep the CSV files so you can see what changed since last time. Before publishing a long article with many sources, check it once too: sources you found months ago may already have moved. If the article is long and dense, the Readability Checker is a useful second pass while you are editing.

Try the free Broken Link Checker

Frequently asked questions

Does a broken link hurt SEO?

A few rarely matter on their own. The clearer costs are a poor visitor experience and, for broken pages on your own site, losing the value of links pointing to them.

Can this check my whole website at once?

No. It checks one page at a time, up to 300 links. Enter your other pages one by one.

Why is a working link marked Blocked?

Some sites refuse automated requests or require a login. Open the link in a normal browser to confirm.

Are redirects a problem?

Not usually. The link still works. Updating it to the final address saves a step for visitors and avoids surprises if the redirect is ever removed.

Does it check images too?

Yes. Images on the page are checked alongside links and marked Image in the results.

Why does a link work for me but show No answer?

The site may have been slow to respond, or it may ignore automated requests. Try the check again later, or open the link yourself.

Sources