Validate an XML sitemap and its index against Google’s rules

How to use the tool

Enter a site address, for example example.com: the tool reads its robots.txt and follows its “Sitemap:” lines, then tries /sitemap.xml if it finds none. Enter a file address instead to check that specific sitemap; the tool still reads the site’s robots.txt to see whether it is declared there.

When the file is a sitemap index, the tool also reads the sitemaps it lists, in order, up to ten files in all. It then requests ten URLs spread across the sitemaps read to see how they respond. A sitemap that is not online yet can be pasted into the second form and analysed in the browser.

What the tool checks

Each check follows a rule written by Google or by the Sitemap protocol, most of them in the form of the errors Search Console shows in its Sitemaps report.³

The tool also flags duplicate URLs, entries without a date and gzip-compressed sitemaps. The Sitemap protocol allows gzip compression, provided the decompressed file stays under 50 MB;⁴ the tool decompresses it and measures that size.

Size limits and the sitemap index

Google limits each sitemap to 50,000 URLs or 50 MB uncompressed, whatever the format.¹ The Sitemap protocol gives the exact value, 52,428,800 bytes, and adds that a compressed file must meet the limit once decompressed.⁴

Beyond that, the sitemap is split into several files linked by an index file. An index can hold up to 50,000 loc tags, and Search Console accepts up to 500 index files per site.² The sitemaps listed must be on the same site as the index, in its folder or lower, unless cross-site submission is set up.² Google’s Sitemaps report flags an index that lists another index as an error.³

The lastmod date: format and reliability

The W3C note on dates allows six forms, from the year alone (2026) to the full date with hours, minutes, seconds and a fraction of a second, such as 2026-10-18T09:30:15.25+02:00.⁵ The time zone, Z or an offset such as +02:00, becomes mandatory as soon as the time is written: Google says so in its description of the “Invalid date” error.³

The format is not enough. Google says it uses the date if it is consistently and verifiably accurate, and expects the date of the page’s last significant update.¹ The Sitemap protocol says the same: the date is when the page was modified, never when the sitemap was generated.⁴ The tool therefore flags a sitemap where every URL carries the same date, and dates later than tomorrow, the one-day margin covering time zones.

What Google ignores: priority and changefreq

Google puts it in one sentence: it ignores priority and changefreq values.¹ The Sitemap protocol still defines them: priority ranges from 0.0 to 1.0 with 0.5 as default, and changefreq takes values such as daily or monthly, which the protocol describes as a mere hint.⁴ The tool counts these tags without treating them as errors: they do not prevent the file from being read.

For more on what Google actually reads in a sitemap, the article XML sitemaps in 2026: what Google really says covers the Search Off the Record episode on the subject.

The URLs listed

Google asks for fully-qualified, absolute URLs, and tries to crawl each URL exactly as written.¹ The Sitemap protocol limits each URL to fewer than 2,048 characters and requires the same protocol and host as the sitemap: a file published on www.example.com cannot list URLs from a subdomain.⁴

The folder matters too. Unless submitted through Search Console, a sitemap only covers its folder and what it contains, which is why Google advises publishing it at the root.¹ A sitemap that groups several sites requires each of them to be verified in Search Console.¹

Finally, Google advises listing the URLs you want to see in its results.¹ A URL that redirects, returns an error or that robots.txt disallows for Googlebot is not one of them; the Sitemaps report even warns when a sitemap contains URLs blocked by robots.txt.³ The tool tests these three cases: robots.txt on every URL checked, HTTP status on the sample of ten.

The tool’s limits

One check reads ten files at most, indexes included, in index order, and checks the first 20,000 entries one by one; the following ones are counted for the 50,000 limit without being checked. Sitemaps left unread are reported in the result. Each file is read up to 50 MB plus one byte, enough to know whether the limit is crossed.

The sample’s HTTP status is the one the tool receives, without following redirects or reading the pages. A site may respond differently to Googlebot, for instance when a firewall blocks unknown bots. RSS and Atom feeds, which Google also accepts as sitemaps,¹ are recognised but not analysed.

The tool accepts 20 checks per 10 minutes per visitor, across all the site’s tools. No address entered and no sitemap is kept, and pasted XML never leaves your browser.

After the check

A clean sitemap does not guarantee indexing: each page must still be indexable. The indexability checker tests one page in detail, and the redirect checker follows the chains spotted in the sample. When Search Console’s Sitemaps report stays in error for no visible reason, that is a diagnosis I carry out in the SEO support I provide to my clients.

Frequently asked questions