What Is a 404 Error?

A 404 is the response your server sends back when someone — a visitor, a crawler, a link from another site — requests a page that isn't there. It's one of the most common things you'll encounter running any website, and also one of the most over-explained: not every 404 is a problem, and the "fix" isn't the same for every one.

What Is a 404 Error, Exactly?

A 404 is an HTTP status code in the 4xx ("client error") class. That classification matters: it means the server understood the request just fine — the problem is that whatever was asked for doesn't exist at that address. The server isn't broken; the URL just doesn't point to anything.

Browsers typically show a generic "404 Not Found" message (or a site's custom error page, if one exists) when this happens. Crawlers see the same status code and record it the same way — a URL that currently has nothing to offer.

What Does "404" Actually Mean?

The number itself doesn't carry hidden meaning beyond its place in the HTTP status code scheme — 4xx codes are client errors, and 404 specifically signals "not found." It's defined the same way across every server, framework, and CMS, which is why a 404 means the same thing whether it comes from a hand-built site or a major platform.

Common Causes

A page was deleted without a redirect set up to point elsewhere.
A page was moved or renamed during a redesign, and internal links weren't updated.
A typo in a link, either one you wrote or one on another site linking to you.
An external site restructured or shut down, breaking a link you had no control over.
A temporary sync issue — rare, but a CDN or cache occasionally serving a stale "not found" state.

Client Error, Not Server Error

It's worth being precise about this distinction: a 404 lives in the 4xx range because the request — the specific URL — is the issue, not the server itself. That's different from a 5xx error, where the server fails to handle an otherwise-valid request. A 404 tells you "there's nothing at this address"; a 500 tells you "something broke while trying to respond." See our HTTP 500 error guide if that's the error you're actually dealing with.

How to Find 404 Errors

A single page: run it through the Broken Links Checker to see every link on that page and its live status, or check one specific URL directly with the HTTP Status Checker.
A whole site: both of those tools can discover URLs from your sitemap.xml and check them in one pass — up to 500 URLs per scan, dependent on your sitemap actually listing your pages.
What Google has actually found: Google Search Console's Coverage report lists 404s Google's own crawlers encountered, which is the most direct signal of what's actually affecting your crawl and index footprint.

How to Fix a 404

There isn't one universal fix — it depends on why the page is gone:

Content was accidentally deleted or the URL changed unintentionally: restore the content, or set up a redirect from the old URL to the new one.
Content was intentionally removed with a genuinely relevant replacement available: redirect to that replacement — a real one, not just the homepage.
Content was intentionally and permanently removed with nothing to replace it: leave it as a 404, or use a 410 if you want to be explicit that it's not coming back (more on that distinction below).
A broken external link on your own page: update it to a working source, an archived version, or remove it — see our guide on finding and fixing broken links for the full workflow.

When a 404 Is Completely Normal

A retired promotion page, a discontinued product with no replacement, an old one-off announcement — these can 404 without it being a problem. That's just a website changing over time. The moment to actually worry is when the 404s are unintentional and accumulating: typos never caught, pages moved without a redirect, a whole section restructured without updating the links pointing into it.

404 Errors and SEO

The blunt version people repeat — "404s hurt your SEO" — isn't quite right. Google's own crawling documentation treats a normal 404 as a useful, expected signal: it tells crawlers a URL is gone and they can stop revisiting it, which is efficient rather than harmful. A handful of 404s on an actively maintained site is routine, not a red flag.

What actually costs you is scale and pattern: a large, uncorrected volume of broken links that visitors and crawlers keep hitting repeatedly signals a site that isn't well maintained, and wastes crawling attention that could go to your current content instead. For the fuller picture on crawl behavior and status codes generally, see our HTTP status codes guide.

Soft 404s: A Different Kind of Problem

A soft 404 isn't a literal HTTP status code — it's a pattern search engines watch for: a page that returns a success status (typically 200) while its content effectively says "this doesn't exist," whether that's an explicit "not found" message, a near-empty page, or a generic placeholder. Google attempts to detect these and treat them as errors regardless of the stated status code, but relying on that detection is worse than just returning the correct status yourself. If a page is genuinely gone, have it say so with a real 404 or 410 rather than a 200.

404 vs. 410: Which Should You Use?

Both tell a visitor or crawler "nothing is here." The difference is how confidently they say it:

404 (Not Found) — the general case. It doesn't assert whether the page might come back.
410 (Gone) — a deliberate, stronger statement that the resource was intentionally and permanently removed.

Use a 410 when you're certain the content is never coming back and want to signal that clearly; a plain 404 is the safer default when you're unsure, or the removal might be temporary.

Custom 404 Pages

A helpful custom 404 page doesn't fix the underlying broken link, but it softens the experience: include a search box, links to popular or recent content, and your normal site navigation, so a visitor who lands there isn't stuck at a dead end. Just make sure the page itself still returns an actual 404 status — a nicely designed error page that responds with 200 is exactly the soft-404 problem described above.

Common Mistakes to Avoid

Redirecting every 404 to the homepage. It's an unrelated destination for whatever the visitor was actually looking for, and search engines can discount redirects that don't lead somewhere genuinely relevant.
Leaving a custom error page returning a 200 instead of a real 404. That's a soft 404 by definition.
Treating every 404 as urgent. Prioritize by where it lives — main navigation and high-traffic pages first, an old forgotten page last.
Never checking at all. 404s accumulate quietly; a periodic check (monthly is reasonable) catches them before they pile up.

Frequently Asked Questions

What is a 404 error?

A 404 error is the HTTP status code a server returns when it can't find the page a client asked for. It's a client-error response (the 4xx class), meaning the server understood the request but has nothing to return at that specific URL — most often because the page was deleted, moved, or the URL was mistyped.

What causes a 404 error?

The most common causes are a page being deleted or moved without a redirect left behind, a typo in the URL (in a link or typed directly), a site restructuring that changed URL paths, or an external site removing or renaming content you'd linked to.

Does a 404 error hurt my SEO?

Not automatically, and not by itself. A 404 for a page that's genuinely gone is a normal, expected response — Google's own crawling guidance treats it as a useful signal to stop revisiting that URL, not a ranking penalty. What actually matters is scale: a large, uncorrected volume of unintentional 404s (typos, un-redirected moves) that visitors and crawlers keep hitting is a real maintenance problem, not routine upkeep.

What's the difference between a 404 and a 410?

A 404 (Not Found) is the general "nothing here" response and doesn't claim to know whether the page might return. A 410 (Gone) is a deliberate, stronger statement: the resource was intentionally and permanently removed. Search engines can act on a 410 a bit more decisively, since it removes the ambiguity a plain 404 leaves open.

What is a soft 404?

A soft 404 is a page that returns a success status code (usually 200) while its actual content says something like "page not found" or shows an empty/near-empty result. Search engines try to detect this pattern and treat the page as an error anyway, but it's a confusing signal worth avoiding — a genuinely missing page should return a real 404 or 410, not a 200 with error-like content.

Should I redirect every 404 to my homepage?

No. A blanket redirect to the homepage sends visitors and crawlers somewhere unrelated to what they were looking for, and search engines can discount redirects that don't lead to a genuinely relevant replacement. Redirect only to a page that actually satisfies the original intent; otherwise, a real 404 (or 410, if the removal is permanent) is the more honest response.

How do I find 404 errors on my site?

Check a single page's links with the Broken Links Checker, verify a specific URL's status with the HTTP Status Checker, or crawl your sitemap with either tool to catch 404s across many pages in one pass. Google Search Console's Coverage report is also a free, direct source of 404s Google has actually encountered while crawling your site.