Broken Link Checker

What breaks, and what it actually costs you

Not every non-200 response is a problem, and treating them all the same wastes time on links that are working perfectly well.

Why links rot even when nobody touches the page

A well-researched article can be entirely correct at publication and full of dead references three years later, without a single edit. Link rot is something that happens to you rather than something you do.

• Publishers restructure. A site moves from /2023/03/post to /blog/post and redirects only the URLs with traffic.
• Companies get acquired. The documentation you cited is folded into the acquirer's site under a different path, or retired.
• Studies move behind a login. The report is still there; the public URL is not.
• Your own migrations. Internal links written against the old URL structure survive the migration and quietly point at nothing.
• Slugs get edited. Someone improves a page title, the CMS updates the slug, and every internal link written by hand breaks.

The last two are the reason to check internal links after any migration or bulk edit, rather than assuming your own site is the safe part.

How to check and fix the links on a page

Six steps. Step five is the one people get wrong: a timeout is not evidence that a link is dead.

1. 01

   Check the page you are about to publish or promote

   This tool reads one page and checks the links on it, so point it at the article you just finished, the page you are about to link a campaign to, or the old post you are refreshing.

2. 02

   Fix the internal 404s first

   An internal broken link is entirely your problem and entirely within your control. It strands whatever authority the link was passing and drops a real visitor onto an error page mid-task.

3. 03

   Decide what each external 404 should become

   Three options: find where the page moved and update the URL, swap in a different source that makes the same point, or remove the link and rewrite the sentence around it. Leaving a dead citation in place is the one option that is always wrong.

4. 04

   Update links that redirect, rather than leaving them

   A link that returns 301 works, but it adds a round trip and it will eventually break when someone tidies up the redirect rules. Point it at the final destination now.

5. 05

   Treat timeouts as unknown rather than broken

   A timeout can mean a slow server, rate limiting, or a bot block on the target. Open it in a browser before deleting the link: plenty of working sites refuse automated requests.

6. 06

   Re-run after the fix

   Confirm every link on the page returns 200. Caching in a CMS means the corrected href sometimes takes a deploy to appear in the rendered HTML.

What a page check finds, and what it cannot

What this tool is good for

Checking a single page before you publish it, after you refresh it, or before you point a campaign at it. Fifty links is more than almost any single article carries, so for one page the coverage is complete.

What it cannot tell you

• Which other pages link to a broken URL. That needs a crawl of the whole site.
• Whether a page is orphaned, meaning nothing links to it at all.
• Links that exist only in JavaScript-rendered navigation.
• Whether a 404 URL used to earn traffic, which is what decides how urgently it matters.

That last point is the real limit. A broken link matters in proportion to how many people hit it, and that number lives in Search Console rather than in the HTML. If the page you are checking redirects rather than 404s, trace the full chain with the redirect chain checker.

What this tool checks

Frequently asked questions