A page goes live that shouldn’t have. A pricing sheet, a staging URL, an old job posting, a client name that was never meant to be public. The first instinct is almost always to block it in robots.txt, and that is the one move that reliably makes the problem worse. Here is what actually works, in the order you should reach for it.
First, Decide What You Actually Want
Every method below answers a different question, and picking the wrong one is why pages come back. There are three:
- The page should disappear from Google right now, and you will sort out the permanent fix afterwards.
- The page should never appear again, and it can be deleted entirely.
- The page should never appear again, but it has to stay reachable for people who have the link.
Answer that first. The rest is mechanical.
Step 1: For Anything Urgent, Use the Removals Tool
In Search Console, open Removals, then the Temporary Removals tab, and request a temporary removal. Google acts on it within about a day, which is far faster than anything that depends on a recrawl.
Two things to understand before you rely on it. First, it is temporary: a successful request lasts roughly six months, after which the page becomes eligible to appear again unless you file another one. Second, it hides the page rather than removing it. Google keeps crawling the URL; it just stops showing it. The tool affects Google Search only, not other search engines and not the page itself, which anyone with the link can still open.
You also get a choice between removing the URL outright and clearing only the snippet, which strips the description text but leaves the result. Use the snippet option when the page is fine and only the preview text is leaking something.
Treat this step as buying time, never as the fix.
Step 2: Make It Permanent, Using the Method That Matches Your Answer
If the page can be deleted, delete it and let the URL return 404 or 410. Google calls this the most secure option, because the content is genuinely gone rather than merely hidden.
If the page must stay reachable, add a noindex rule. Either a meta tag in the head:
<meta name="robots" content="noindex">
or, for non-HTML files like PDFs, an HTTP response header:
X-Robots-Tag: noindex
If it should only be visible to specific people, put it behind a password. That blocks Googlebot and every other crawler at once, and unlike noindex it does not rely on a search engine choosing to honour a rule.
The robots.txt Trap
This is the part that catches nearly everyone. Blocking a page in robots.txt does not remove it from Google, and it stops the removal from ever happening. Google’s own documentation is direct about it: for a noindex rule to work, the page “must not be blocked by a robots.txt file, and it has to be otherwise accessible to the crawler.” If Googlebot cannot fetch the page, it never sees the noindex, and the URL can keep appearing in results indefinitely.
So the sequence matters. Add the noindex first, confirm the page is still crawlable, and only consider a robots.txt rule much later, once the URL has actually dropped out. In practice there is rarely a reason to add one at all.
What Trips People Up
Four details worth knowing before you file anything:
- A noindex needs a recrawl to take effect, and Google may not revisit a low-priority page for months. Use the URL Inspection tool to request indexing and push it along.
- If the URL returns a 404 during a temporary removal, the block expires early. Google then treats the URL as new content if it ever comes back. Deleting the page and relying on the Removals tool at the same time can therefore backfire.
- Prefix removals match more than you might think. Blocking a prefix covers http and https, and www and non-www, and everything nested beneath it. It does not cover other subdomains such as
m.oramp., which have to be requested separately. - You need to own the property. If the page is on a site you do not control, the Removals tool is not available to you. The Outdated Content tab handles pages that have already been changed or deleted, and there is a separate process for personal information.
The Check Afterwards
A week later, search site: plus the exact URL, and inspect it in Search Console. If it still shows, the usual cause is the one above: something is blocking the crawl, so the rule you added has never been read.
Knowing which lever to pull is most of technical SEO in miniature, and the same diagnostic habit applies when pages you do want indexed are misbehaving, which is the subject of reading Search Console after a traffic drop. If something is indexed that shouldn’t be and it isn’t going away, get in touch.