A Google site: query is useful for spot checks, but it does not necessarily return every indexed URL and should not be used as an exact index count. Use URL Inspection to verify one specific page and the Page Indexing report to evaluate sitewide indexed and non-indexed patterns.
- Absence from a site: query does not prove that a URL is not indexed.
- Broad site: result counts are not an index census.
- Use exact URL Inspection evidence for important pages and narrower site: queries only as supporting clues.
01
Identified claim
A common claim says:
“The number shown by
site:example.comis the number of pages Google has indexed.”
That claim is false.
The site: operator limits search results to a domain, URL, or URL prefix. It can reveal examples of pages Google can serve, but Google says the list is not necessarily exhaustive. How to use the site search operator[1]
Use it as a debugging clue, not an index database.
What the operator does
Examples:
site:example.com
site:https://example.com/articles/
site:example.com canonical tags
site:https://example.com/articles/example-page/The operator changes the scope of a normal search. It still uses Google’s retrieval and serving systems.
Google documents that:
- indexed URLs can appear in relevant
site:searches; - appearance is not guaranteed;
- broad results may omit indexed URLs;
- large sites should not expect an exhaustive list;
- result order is not an index order or ranking report.
02
Sources and evidence
Why the result count is unreliable
A broad query such as:
site:example.comdoes not request a formal database count. The displayed estimate can vary as Google retrieves results.
Do not use it to measure:
- exact index size;
- day-to-day indexing changes;
- canonical inventory;
- whether every sitemap URL is indexed;
- whether a specific missing page has a technical defect.
Search operators are bounded by indexing and retrieval limits. Google states that URL Inspection is more reliable for debugging a specific page. Debugging with Google Search operators[2]
Absence does not prove non-indexing
A URL can fail to appear in one site query because:
- the query is broad;
- the URL is not selected for that result set;
- the prefix is wrong;
- HTTP, HTTPS,
www, and non-wwwdiffer; - the query includes terms the page is not relevant to;
- another canonical represents the duplicate cluster;
- serving conditions differ.
Use URL Inspection for the exact canonical URL.
The tool reports what Google knows about one page and can test the live version. URL Inspection Tool[3]
Appearance does not prove a healthy page
A URL can appear in a site: query and still have problems.
Examples:
- indexed though blocked by robots.txt;
- URL-only result with limited snippet;
- stale title or cached information;
- alternate URL served for a narrow query;
- indexed page with weak ranking elsewhere;
- a page Google later drops after another crawl.
The operator shows that Google served a result. It does not validate every technical field.
Use the right tool for the question
| Question | Best starting tool |
|---|---|
| Is this exact canonical URL indexed? | URL Inspection |
| Why is this exact URL not indexed? | URL Inspection and live test |
| How many known URLs are indexed or excluded by reason? | Page Indexing report |
| Which queries produce impressions and clicks? | Search Performance |
| Can I find examples of indexed URLs under a prefix? | site: operator |
| Are unexpected spam URLs appearing? | Targeted site: queries |
| Did a recent fix reach the live page? | HTTP inspection and live URL test |
The Page Indexing report provides Google’s aggregate indexed and non-indexed totals for known pages, while example listings remain limited. Page indexing report[4]
Useful site queries
The operator remains useful when used narrowly.
Find a specific page
site:https://example.com/articles/example-page/This may reveal whether the page is being served, but URL Inspection remains authoritative for diagnosis.
Inspect a section
site:https://example.com/articles/Use this to sample which article URLs appear, not to count the complete section.
Find suspicious pages
site:example.com casino
site:example.com viagraThis can expose spam, hacked pages, or indexed garbage.
Check topic association
site:example.com canonicalizationThis can show which pages Google serves for a topic and may reveal internal cannibalization.
Compare hostname or protocol variants
site:https://example.com/
site:https://www.example.com/Treat differences as clues. Verify actual canonical and redirect behavior separately.
Common mistakes
| Mistake | Why it fails |
|---|---|
| Reporting the displayed result estimate as exact index count | It is not a complete census |
| Declaring a page deindexed because one query omits it | Results are query-dependent |
| Checking the wrong URL prefix | The operator scopes literally |
| Ignoring canonicalization | Another URL may represent the content |
| Using site search instead of URL Inspection | The query lacks URL-level diagnostics |
| Tracking daily site counts | Normal result variation looks like index volatility |
| Assuming first result is the strongest page | Broad site queries are not ordinary rankings |
03
Conclusion
The claim that site: produces a complete list or exact count of indexed pages is false.
Use this operating rule:
Use
site:to sample what Google can serve, URL Inspection to diagnose one URL, and the Page Indexing report to evaluate aggregate coverage.
For a publication batch:
Verify sitemap and canonical routes. Inspect a representative set of exact URLs. Monitor Page Indexing reasons and report dates. Use Search Performance for impressions and queries. Use site: only as an additional spot check.
This avoids turning an approximate search view into a fictitious analytics system.
Broad site: result counts are not an index census.
04
Limitations
URL Inspection itself has boundaries. The live test shows current eligibility, not guaranteed indexing, and indexed information can reflect an earlier crawl.
The Page Indexing report can lag and limits examples even though its totals represent Google’s known inventory.
No public tool exposes Google’s entire internal index as a downloadable list.
The correct answer therefore depends on the question. site: remains useful, but it is the wrong instrument for exact index accounting.
Absence from a site: query does not prove that a URL is not indexed.
References
Sources behind this record
- How to use the site search operator — Google (accessed July 26, 2026)
- Debugging with Google Search operators — Google (accessed July 26, 2026)
- URL Inspection Tool — Google (accessed July 26, 2026)
- Page indexing report — Google (accessed July 26, 2026)
Corrections
Correction history
No corrections recorded.
To report an error, use the public corrections path.
Search results vary by query and serving conditions, so a site: query cannot reproduce Google’s complete internal index. No public tool exposes Google’s entire internal index as a downloadable list.