Check the links on a page
Paste a page address. We request every link on it and show you which ones are broken, which ones answer, and which ones would not tell us anything.
Checking the links
Loading the page…
We are requesting the links one at a time and waiting for each server to answer. Expect around thirty seconds, longer if a site is slow.
This is taking longer than usual: one server is slow to answer. We give up on it after eight seconds and move on to the next.
How to read this result
The right-hand column shows three verdicts, not two. The distinction matters: most of the “broken link” reports going around flag links that work perfectly well.
Dead
The server answered with an error, usually 404 (the page does not exist) or 410 (it was deliberately removed). This is the only case where there is really something to do: fix the address, replace it, or drop the link.
Blocks bots
The server answered 403, 429 or 503. That tells you nothing about the page: it tells you a firewall pushed us aside. Plenty of large sites turn away automated requests on principle, and they are often the ones most worth citing. Removing those links means deleting links that open perfectly for your readers.
No answer
Nothing came back within the eight seconds allowed. A slow server is not a dead server, and a three-second network hiccup is not a vanished page. Re-check these later before touching anything.
Internal links, external links
A broken link on your own site can be fixed right now: it is yours, you have the keys. A broken link on someone else's cannot, the page at the far end is gone. It is still worth knowing, because an article riddled with dead links ages badly, for your readers and for Google alike.
What this tool does not do
- It looks at one page, and twenty-five links at most. A whole site needs a crawl, not twenty passes over this form.
- It only sees links present in the served HTML. Anything a script adds afterwards escapes it.
- It ranks nothing. A broken link in a footer and one in the middle of an article do not carry the same weight, and that call is yours.
Frequently asked questions
A link that works is marked “blocks bots”. Why?
Because the site guards itself against robots, and we are one. Its firewall answers 403 to any automated request, whoever sends it. The link opens normally in a browser, and we would rather say so than wrongly condemn it.
How long does the check take?
From a few seconds to about a minute, depending on the number of links and the speed of the servers we contact. Each link gets eight seconds to answer; past that we move on and mark it “no answer”.
Why only twenty-five links?
Because every link tested leaves our servers for a site we know nothing about. Without a cap, this free page would become a convenient way to fire thousands of requests at any site. The cap protects the sites you test.
Are nofollow links tested?
Yes. The rel attribute tells search engines not to follow the link; it says nothing about the state of the page at the far end. A nofollow link to a dead page is still a dead link for your readers.
Going further
Find broken links by hand, without a paid tool Four free ways to find the broken links on a site, what each one misses, and the point where doing it by hand stops being reasonable.The other tools
Check a domain name
Expiry date, registrar and status of a domain name.
Check an SSL certificate
Certificate expiry date, issuer, and the names it actually covers.
Check redirects
The whole redirect chain, hop by hop, with the type of each one.
One page by hand, or the whole site on its own
Expansel walks your pages, tests the links on each one and hands you the list of what is broken, with the page where each dead link was found. One page by hand here, the whole site in a single crawl there.