To remove a URL on your own site from Google you need two steps: the Removals tool in Search Console, which usually hides the page within a day for about six months, and a permanent fix on the site (a 404 or 410 status, a noindex tag or a password). If the page belongs to someone else, Search Console won't help: you use Google's outdated content tool or its personal information requests instead.
Put like that, it's a five-minute job.
Then you search for how to do it, and the answers disagree.
On 28 October 2026 we read the first page of 8 Italian searches on the topic: of 70 organic results, 18 came from law firms and online reputation companies, 15 from SEO agencies and consultants, and 8 were about something else entirely (clearing Chrome history, deleting pages from a PDF).
The AI Overview, present on all 8, contradicted itself: in 4 answers the temporary removal lasted "about 6 months", in 2 "about 90 days".
Which is it?
Six months. Google's help page for the Removals tool says it blocks the URL from Google Search results "for about six months" (checked on 28 October 2026). The UK results for "remove url from google" got this right: the AI Overview said "approx. 6 months", and none of the 7 organic results came from a law firm.
When a company asks us to make a page disappear, we nearly always find one of these three mistakes:
- using only the temporary removal, and finding the page back in the results six months later;
- blocking the URL in robots.txt, which stops Google reading the noindex;
- using the wrong tool for a page they don't own, and getting the request denied.
This article comes from our own site.
Between September and October 2026 we changed the CMS and the language plugin on visilay.com (the full story is in our CMS migration case), and along the way some pages were removed or moved. We followed them in Search Console: 9 URLs that now return a redirect or a 404 collected 434 impressions and 1 click between 1 and 26 October.
Here's what you'll find:
- who answers on Google when people search how to remove a URL;
- temporary vs permanent removal, in Google's own words;
- which route to take in six situations, including pages you don't own;
- how long a removed URL stays visible, with visilay.com data;
- what changes for AI Overviews, Bing and ChatGPT.
Who answers on Google when you search how to remove a URL (8 Italian searches, 70 results)
On 8 Italian searches about removing or deleting a URL from Google, the most frequent source on page one was Google itself (22 of 70 results, mostly Help Community threads), followed by law firms and reputation companies (18) and SEO agencies and consultants (15). The page that actually explains the procedure, Google Search Central's documentation, never appeared among the organic results: it appeared as a source inside the AI Overview.
| Italian search (meaning) | Duration given by the AI Overview | Law firms and reputation | SEO agencies and consultants | Google pages | Off topic |
|---|---|---|---|---|---|
| rimuovere url da google (remove url from google) | about 6 months | 3 | 2 | 2 | 1 |
| rimozione url google (google url removal) | about 6 months | 3 | 1 | 3 | 0 |
| rimuovere pagina da google (remove page from google) | about 90 days | 1 | 3 | 2 | 3 |
| deindicizzare pagina google (deindex page google) | not returned | 3 | 2 | 3 | 0 |
| rimuovere link da google (remove link from google) | about 6 months | 2 | 4 | 1 | 0 |
| eliminare url da google (delete url from google) | about 6 months | 4 | 1 | 0 | 3 |
| come rimuovere un sito da google (how to remove a website from google) | about 90 days | 2 | 1 | 3 | 1 |
| rimuovere contenuti obsoleti google (remove outdated content google) | not returned | 0 | 1 | 8 | 0 |
| Total (70 results) | 4 of 6 say 6 months | 18 | 15 | 22 | 8 |
Three things stand out.
Two intents behind one search. A site owner wants to remove their own page. Someone who finds their name in an article wants to remove someone else's. In Italy, law firms answer the second with the right to be forgotten and GDPR, and the site owner reads an answer that isn't about them.
The wrong duration travels through guides. Among the organic results there is at least one 2025 SEO guide giving a removal "valid for 90 days", and the two AI Overviews that say 90 days cite hosting and agency guides, not Google Search Central.
Nobody shows what happens next. Every guide tells you where to click. None says for how many days a removed URL keeps showing up, or whether it matters. That's the part we show you with our own data.
Temporary vs permanent removal (and why you need both)
A temporary removal hides a URL from Google's results without touching the site; a permanent removal changes the page or the server response, and Google picks it up on the next crawl. The first is fast and expires. The second is slow and sticks.
This is what Google's documentation says about the Removals tool in Google Search Console, read on 28 October 2026:
- it lasts about six months, then the page can reappear unless you've made the removal permanent;
- it usually takes up to a day, and the request "is not guaranteed to be accepted";
- it doesn't stop crawling: Googlebot keeps visiting the page, it just doesn't show it;
- it only works on Google Search: other search engines know nothing about it;
- you must own the Search Console property the URL belongs to.
For the permanent part, Google Search Central's page "Remove a page hosted on your site from Google" (updated 31 December 2025) lists three routes: remove or update the content, password-protect it, or add noindex. And one thing not to do: use robots.txt as a blocking mechanism.
So what's the temporary removal for, if you have to fix the site anyway?
It covers the wait. A noindex or a 404 only works once Googlebot comes back, and on a page with few links that can take weeks. If the page shows a confidential price list or a client's data, those weeks are too long: request the temporary removal today and fix the site in the meantime.
The full comparison of noindex, Disallow, canonical, 404 and passwords is in our noindex guide, with a table of what blocks crawling and what removes from the index. Here we focus on the practical choice.
Which route to take (six situations and Google's answer)
The right route depends on two questions: does the page need to stay online, and is it yours? With those two answers there's almost always only one choice.
1. The page should no longer exist
Delete it and let the server return a 404 or 410. In its documentation on HTTP status codes Google says that already indexed URLs returning a 4xx "are removed from the index" and that all 4xx errors except 429 are treated the same. A 410 is not a fast lane. When to use which is covered in our article on the 404 error.
2. The page stays online but out of Google
Thank-you pages, client areas, campaign landing pages: this is a job for noindex, in the meta tag or the HTTP header. It only works if Googlebot can read it, so the URL must not be blocked in robots.txt. How to write it and why it sometimes fails is in our noindex guide.
3. The page has a replacement
A renamed service, a merged article, a replaced product: 301 redirect to the new page. Google treats a 301 as a "strong" signal that the target should be processed. It isn't an instant removal: as our data shows, the old URL stays visible for a few weeks.
4. You published something confidential by mistake
This is where the temporary removal earns its keep. First password-protect or remove the page, then request removal in Search Console: the page drops out of the results in about a day, and when the request expires Google will find the page already closed. If the results also showed a snippet with the data, the option to clear the snippet removes it until the page is indexed again.
5. You need to remove many URLs at once
A folder of old PDFs, pages created by a hack, a staging site that got indexed. The Removals tool accepts a prefix ("Remove all URLs with this prefix") and on its own covers http and https, www and non-www. Subdomains don't match: Google says m. or amp. need separate requests. For PDFs see our piece on PDF SEO; if the pages come from a hack, first check whether Google has taken action, as we explain in website banned from Google.
6. The page isn't yours
Search Console doesn't apply: the Removals tool only works on URLs in a property you own. There are three routes, in this order.
- Ask the site owner to remove or change the page. It's the only route that removes the content from the web, not just from Google.
- If the owner has already removed or changed it but Google still shows the old version, use the Refresh Outdated Content tool: you don't need to own the site.
- If the page is live and shows your contact details, Google has a process to find and remove personal contact info from results.
Right to be forgotten requests under UK GDPR or EU GDPR are a legal matter: we're not lawyers, which is why that part of the Italian results belongs to law firms rather than SEO agencies.
How long a removed URL stays on Google (visilay.com data)
A URL removed from a site stays in the results until Google recrawls it: on the two visilay.com service pages we binned on 8 October 2026, the last impression came 16 days later for the English version and 18 days later for the Italian one. With no request in Search Console.
The background. While we switched language plugins, some new Italian articles were also indexed at the address without /it/, which is reserved for English pages. Those addresses now return a 301 to the Italian version, two of them a 404. In the same period we closed the "SEO AI" service page and merged it into our GEO service.
| URL | What happened | Response on 28 October | Impressions 1-26 October | Last day in results |
|---|---|---|---|---|
| /services/seo-ai/ | service page binned on 8 October | 301 to /services/geo/ | 42 | 24 October |
| /it/servizi/seo-ai/ | service page binned on 8 October | 301 to /it/servizi/geo/ | 4 | 26 October |
| /blog/migliori-servizi-seo/ | Italian article indexed without /it/ | 301 to the /it/ version | 283 | 26 October |
| /blog/costi-local-seo/ | same case | 301 to the /it/ version | 58 (1 click) | 26 October |
| /blog/statistiche-motori-di-ricerca/ | same case | 301 to the /it/ version | 19 | 26 October |
| /blog/sfide-seo/ | same case | 301 to the /it/ version | 12 | 25 October |
| /blog/sito-non-genera-lead/ | same case | 301 to the /it/ version | 6 | 15 October |
| /blog/migliori-newsletter-seo/ | same case | 404 | 7 | 3 October |
| /blog/miglior-crm-per-seo/ | same case | 404 | 3 | 24 October |
| Total 9 URLs | 7 redirects, 2 404s | 434 (1 click) |
434 impressions and 1 click. Here's what that tells us.
A redirect doesn't remove a URL overnight. The English "SEO AI" page picked up 32 impressions after it was binned, on 10 different days. The old address kept showing until Google came back and followed the 301.
Not every removed URL needs a manual request. These addresses brought one click in 26 days. A temporary removal would have hidden them a few days earlier without changing anything for the site: the real work was the redirect.
Two mistakes are ours. The two URLs returning a 404 are duplicates of articles that exist: the right answer would have been a 301 to the Italian version, as with the other five. A 404 on a duplicate does no serious harm, but it throws away the signals that address had collected.
And if those URLs had shown a confidential price list?
Then 18 days would have been far too long, and we'd have requested the temporary removal, which works in about a day, the same morning.
Old pages, duplicates or URLs that should never have reached Google?
You've just seen 9 removed addresses that stayed in the results for up to 18 days, and two duplicates that ended in a 404 instead of a 301. Using Search Console and server data we build the list of URLs to remove with you, decide 404, noindex or redirect for each one and check they actually drop out.
Book a strategy callThe mistakes that keep a URL on Google longer than it should be (four, all documented)
A URL stays in the results longer than necessary almost always because something stops Google from seeing the change. These are the four Google's documentation describes explicitly.
- Disallow in robots.txt instead of noindex. Robots.txt blocks crawling, not indexing: the URL can stay in the results without a description, and Google will never read the noindex on the page.
- Temporary removal with no fix on the site. After about six months the request expires and, if the page still returns 200 without noindex, it can come back.
- Forgotten subdomains and variants. The tool covers http, https, www and non-www, but not m. or amp.; and a URL with different parameters or capitals is a different URL, as Google reminds us with /puppies and /PUPPIES.
- Canonical used as a deletion. A canonical consolidates signals onto another address, it removes nothing from the index. It's in the table in our noindex guide, and it's the mistake that leaves you waiting for an effect that never comes.
There's also a check that costs nothing: the Page indexing report and the URL Inspection tool, which shows what Googlebot saw last time. If the page is still "indexed" a week after the 404, the issue is usually crawling, and our article on Google indexing explains how to read the statuses.
If you're removing many pages to clean up the site, the question to ask first is which ones to remove: we covered it, with numbers, in our piece on content pruning.
Does removing a URL from Google also remove it from AI answers? (it depends on the engine)
From AI Overviews and AI Mode, yes, once the page leaves the index: to appear as a supporting link a page must be "indexed and eligible to be shown in Google Search with a snippet", Google writes on its AI features page (updated 10 December 2025). A page returning 404 or carrying noindex doesn't qualify.
From other engines, not with Google's tool. The Removals help page is explicit: it removes content only from Google Search. Bing has its own index and its own tools, which we cover in our article on Bing Webmaster Tools and IndexNow; assistants such as ChatGPT read the web with their own crawlers, listed in our guide to AI crawlers. A 404 works for any crawler, noindex for those that support it (Google itself warns some engines may ignore it); a Search Console request only works for Google.
There's a flip side. In the 6 readable AI Overviews from our Italian check, Google Search Central's documentation was cited in 4. The 2 that said "90 days" didn't cite it: their sources were hosting and agency guides. A clear, up-to-date official page is what ends up in the answer; where it's missing, the answer takes someone else's guide, mistakes included. If an AI answer gets facts about your company wrong, that's a different route, covered in our piece on wrong information in AI Overviews.
When removal is part of a bigger job, such as a platform change or cleaning up years of accumulated pages, it's the kind of work we handle in our SEO services.
FAQs about removing a URL from Google
No. Google's help page for the Removals tool says about six months. The 90-day figure appears in some guides and in 2 of the 6 AI Overviews we read on Italian results on 28 October 2026, but not in Google's documentation.
It depends: not with the Removals tool, which requires owning the site in Search Console. If the owner has already removed or changed the page you can use the Refresh Outdated Content tool; if the page shows your contact details there is Google's personal information process; right to be forgotten requests need a lawyer.
No. Robots.txt blocks crawling, not indexing: the URL can stay in the results without a description, and Google can't see the noindex on the page. Google itself advises against using robots.txt as a blocking mechanism.
No. In its documentation on HTTP status codes Google says all 4xx errors except 429 are treated the same: already indexed URLs are removed from the index when Googlebot recrawls them.
No. Google says the tool removes content only from Google Search. For other engines the page response is what counts: a 404 works for any crawler, noindex for those that support it.
No, usually not. On visilay.com redirected URLs stayed visible for up to 18 days and brought 1 click in 26 days: Google replaces them on its own when it recrawls. Removal is worth it only if the old address shows something that must not be seen.
Removing a URL from Google, in short
Your page, gone for good: 404 or 410.
Online but out of Google: noindex, no Disallow.
Page with a replacement: 301, and a few weeks of patience.
Emergency: temporary removal, about six months, plus the fix on the site.
Someone else's page: site owner, outdated content tool, personal information request.
The Search Console removal is the most searched part and the one you need least often.
The check we'd suggest is the one we used to find the 9 URLs in this article: export from Search Console the pages with impressions over the last few weeks and check the HTTP response of each one. The ones returning 301 or 404 that still get impressions show where Google hasn't caught up with a change, and which duplicates were created by accident. It's a short list, and that's usually where the work is, well before asking Google to hide anything.
Speak soon!
