Server response codes and SEO: what 200, 3xx, 404, 410, and 5xx mean

How server response codes are structured

Five classes of codes

Codes fall into five classes based on the first digit:

Class	Range	Meaning
1xx	100–199	Informational — an intermediate response, the request is still being processed
2xx	200–299	Successful — the request was completed
3xx	300–399	Redirects — further action is needed, usually moving to another address
4xx	400–499	Client errors — the problem is in the request (for example, the page doesn’t exist)
5xx	500–599	Server errors — the server couldn’t process a request that was itself valid

How to see a page’s response code

The response code is visible in the browser’s developer tools, on the network requests tab. Separately, Yandex Webmaster has a «Server response check» tool for your own site: it shows whether pages are accessible to Yandex’s crawlers, along with other diagnostic information. If the server’s response isn’t a 200 OK code, the page can’t appear in search results.

2xx and 3xx: the page is available or has moved

A 200 OK code means the request was completed successfully — this is the normal response for a live page.

3xx codes are redirects: the server reports that the resource should be looked up at another address. The main differences between 301 (a permanent move) and 302 (a temporary one) are a separate topic with setup examples; here it’s only important to know that a 3xx code isn’t itself an error — it’s an instruction to go to a new address.

Code	Class	What it means
200	2xx	The request was completed successfully
3xx	Redirect	The resource should be looked up at another address
404	4xx	The resource was not found
410	4xx	The resource has been permanently removed
429	4xx	Too many requests
5xx	Server error	The server couldn’t process the request

4xx: client errors

404 Not Found

A 404 code means the server didn’t find a current representation of the requested resource. Yandex Webmaster explains it as: the document doesn’t exist. If a page was removed without a replacement, 404 is a normal outcome and isn’t a problem for the site overall. If the page has moved instead, you need a redirect, not a 404.

Google treats all 4xx errors except 429 the same way: it tells its processing systems that the content is missing. Discovered pages with a 404 code aren’t processed, and the frequency of retrying to crawl them gradually decreases.

410 Gone

A 410 code means access to the resource is no longer provided on the origin server, and this is most likely a permanent condition. Yandex puts it as «the resource has been permanently removed from the site» — unlike 404, where the reason for the absence isn’t specified. In practice, Google treats 410 the same as 404 when crawling, but for you as a site owner the difference is meaningful: 410 is a deliberate signal that «there will be no more content here,» while 404 can mean either a temporary error or a typo in the address.

429 Too Many Requests

A 429 code means the client sent too many requests in a limited time. For a site, that’s a signal to set up rate limiting if the load is causing problems.

Soft 404

A soft 404 is when a nonexistent page returns a code other than 404 or 410, or a redirect to another page (for example, the homepage) happens instead of a 404 code. An example is an empty on-site search results page with a «nothing found» message that still responds with a 200 code. Google records such cases in Search Console as a soft 404 error. The problem is that both the user and the search engine get a contradictory signal: the page looks like an error, but it’s formally marked as working.

	404	410	Soft 404
Response code	404	410	200 (or a redirect) while the content is actually missing
Meaning	The resource wasn’t found, the reason isn’t specified	The resource has been permanently removed	The code doesn’t match the page’s actual content
How Google treats it	The content is missing; the crawl frequency decreases	Treated the same as 404	Flagged as a separate error in Search Console

5xx: server errors

5xx codes mean the server ran into a problem while processing the request, even though the request itself was valid:

• 500 Internal Server Error — the server ran into a situation it couldn’t handle and found no more fitting 5xx code to respond with.
• 502 Bad Gateway — a server acting as a gateway or proxy received an invalid response from another server.
• 503 Service Unavailable — the server isn’t ready to handle the request; a common cause is maintenance or overload.
• 504 Gateway Timeout — a server acting as a gateway didn’t get a response from an upstream server within the allotted time.

According to Google, during 5xx server errors crawlers temporarily slow down crawling: already-indexed URLs stay in the index but are removed if the errors continue; once the server starts responding with a 2xx code, the crawl rate gradually recovers. Yandex Webmaster’s HTTP code reference doesn’t describe how codes 500, 502, 503 and 504 affect indexing or crawling — it only gives technical recommendations for fixing the problem (checking logs, server capacity, settings).

Summary table: server response codes and SEO

Code	What it means	Google	Yandex
200	The request was completed successfully	The page is indexed	The page is indexed normally
3xx	A redirect to another address	Handled per redirect rules	Handled per redirect rules
404	The resource was not found	Content is missing; retry frequency decreases	The document doesn’t exist
410	The resource has been permanently removed	Treated the same as 404	Used for content that has been permanently removed
429	Too many requests	—	Set up rate limiting
500 / 502 / 504	Server error	Temporarily slows crawling; indexed URLs stay but are removed if the errors continue	The reference doesn’t detail the effect on indexing
503	The service is temporarily unavailable	Crawl rate recovers once a 2xx code returns	The reference doesn’t detail the effect on indexing

How to set up a 404 page

• The page should actually respond with a 404 code (or 410, if the removal is permanent), not a 200 code and not a redirect to the homepage — otherwise it’s a soft 404.
• A 404 page doesn’t have to be empty: you can add navigation on it so the user doesn’t lose their way around the site.
• A 404 page shouldn’t be blocked in robots.txt without a reason — otherwise it’s harder for the crawler to confirm that it’s really an error page. More on how the robots.txt file works — in a separate article.
• If a page has moved rather than been permanently removed, you need a redirect, not an error, instead of 404.

How to check a page’s response code

• Browser developer tools. The network requests tab shows the response code for any open page.
• Yandex Webmaster — «Server response check.» Shows the response code and the SSL certificate status for a specific address.
• Google Search Console — the page indexing report. A «Not found (404)» status means the request for the page got a 404 code back; Google notes that the crawler will most likely keep trying to find the URL, but will do so less and less often.

If the response code is fine but the page still doesn’t make it into the index, the reason may lie elsewhere — more on how pages enter and drop out of the index, in the article on indexing.

Frequently asked questions

What’s the difference between 404 and 410?

404 means the resource wasn’t found, without specifying the reason. 410 means the resource has been permanently removed — it’s a deliberate signal, not an error.

Does a large number of 404 pages hurt a site?

Usually not, as long as those addresses genuinely shouldn’t exist on the site.

What should you do if a page has moved — return 404 or set up a redirect?

A redirect, if there’s a suitable new address. 404 or 410 — if the content is gone for good and there’s no replacement.

What is a soft 404 and how do you avoid it?

It’s when a missing page returns a 200 code or a redirect instead of a genuine 404/410. To avoid it, set up missing content to respond with a 404 or 410 code, not 200.

How long does Google keep retrying to crawl a page with a 404?

The crawler keeps trying, but does so less and less often over time.

Does a 503 block a site from being indexed?

Not directly, but with a 503 (as with other 5xx errors) Google temporarily slows down crawling the page; the rate recovers once the server starts responding with a 2xx code again. Yandex Webmaster’s reference for this code doesn’t separately describe the effect on indexing.