How to Fix Broken Links: Redirect, Replace, or Remove

August 23, 2026 · 8 min read · Broken Link Checker team

Every broken link has one of three correct fixes: redirect the dead URL to its living successor, replace the link at the source with a working one, or remove it because nothing serves its purpose anymore. The skill is picking the right one — a redirect where a source fix belongs creates chains, and a homepage redirect where a 404 belongs creates soft 404s. This guide is the decision tree, plus the traps.

Step 0: triage before you touch anything

For each broken link you found (if you haven't yet, start with how to find broken links), answer two questions:

Whose URL is dead? If it's on your domain, you control both ends — you can revive the page or redirect it. If it's on someone else's domain, you can only change your link.
Was it really broken? Check the status. A 404/410 is dead; a 5xx might be a temporary outage worth rechecking tomorrow; a 403/429 is often a site blocking bots while working fine for humans. A checker that labels these separately — as Broken Link Checker does with its "blocked" category — saves you from "fixing" links that were never broken. (Reference: what each status code means.)

Fixing broken links on your own site

Case 1: the URL changed — redirect it

Content moved, slug renamed, site migrated. Set a 301 (permanent) redirect from the old URL to the new one. This repairs every copy of the link at once — bookmarks, other sites' links, search results — and passes the old URL's accumulated authority to the new address.

Rules that keep redirects healthy:

Redirect to the closest equivalent, not the homepage. A homepage redirect ignores what the visitor asked for; Google classifies it as a soft 404 and treats it like the error it is.
Use 301, not 302, for permanent moves. A 302 says "temporary," which asks search engines to keep the old URL indexed.
Avoid chains. If A→B and later B→C, edit the first rule to A→C. Every hop adds latency, and long chains get abandoned by crawlers.
Keep a redirect map during migrations: a spreadsheet of old → new for every URL, written before the switch. Migrations without one are where most of a site's authority quietly dies.

Case 2: the link itself is wrong — fix it at the source

A typo, a pasted URL that swallowed a trailing character, an extra https://. Don't redirect around a typo — correct the link in the page or CMS. Redirects are for URLs you can't reach into (other people's links); your own links should point straight at the real destination.

Even when a redirect exists, update internal links to skip it: internal links that go directly to the final URL are faster for visitors and cleaner for crawlers.

Case 3: the content is gone on purpose — let it 404 properly

Discontinued product, retired campaign, deleted post with no successor. The correct behavior is an honest 404 — or 410 Gone if you want search engines to drop it faster. Then:

Remove your own internal links pointing at it (they're now knowingly broken).
Make the 404 page useful: say plainly the page doesn't exist, offer search and a few paths back in. A good error page turns a dead end into a detour.

Fixing broken links to other sites

When an external page you linked to dies, you have three options in order of preference:

Find the content's new home. Sites reorganize; the article often still exists at a new URL. Search for its exact title — if it moved, link the new address.
Link an archived copy. The Wayback Machine at web.archive.org stores snapshots of most public pages. Paste the dead URL there, pick a snapshot from when you originally linked it, and use the archive URL. The reference survives even though the site didn't.
Rewrite or remove. If nothing equivalent exists, unlink the text or cite a different source. A sentence without a link beats a link to nowhere.

Tip: when you fix external links on a page, rescan the whole page afterwards — pages old enough to have one dead reference usually have several. One scan with the extension shows them all at once, with a CSV export if you want a to-do list.

Reclaiming links other sites point at you

The reverse case is the profitable one: someone else's site links to a URL of yours that now 404s. You can't edit their page, but a 301 redirect on your side reconnects their visitors — and the link's SEO value — to living content. Find these URLs in Google Search Console (Indexing → Pages → "Not found (404)") and redirect any with real inbound links. The ranking mechanics are covered in do broken links hurt SEO?

The mistakes that make things worse

Mistake	Why it backfires
Redirecting every 404 to the homepage	Soft 404s in Google's eyes; disoriented visitors in everyone else's.
Deleting pages "to clean up 404 reports"	The report was never the problem. If the URL has inbound links, deleting harder loses more.
Stacking redirects over the years	Chains slow every visit and eventually stop passing authority. Point the oldest rule at the final URL.
"Fixing" bot-blocked links	A 403 from a bot-hostile site isn't broken for humans. Verify in a normal tab before replacing a perfectly good source.
Fixing once, never rechecking	External links keep dying at their own pace — link rot never stops. Put a periodic scan on the calendar.

A workflow that takes minutes, not afternoons

Open the page in Chrome and run Broken Link Checker — every link is tested in seconds.
Work the red list: your URLs → redirect or restore; external URLs → replace with live or archived versions; typos → correct in place.
Ignore the "blocked" list unless a spot-check in a normal tab actually fails.
Rescan. Zero problems? Ship it.
Repeat monthly for key pages, and after every migration for everything.