Sitemap validator that names the rule each problem breaks

01Validate a sitemap

Paste the XML. Adding the URL the sitemap is served from is optional and switches on two extra checks, whether any entry points at a different host and whether the file is where it claims to be. Nothing is uploaded and the page makes no network request while you use it.

Validation uses the browser's own XML parser, which is the same engine that decides whether the file is well formed in the first place. A sitemap that fails here fails for a reason the parser states, not for a reason this page invented.

02Two fields the spec allows and Google throws away

The sitemaps.org schema defines four child elements for a URL entry. Google reads two of them.

Google's documentation is unambiguous about the other two. It states that it ignores changefreq and priority, and that lastmod is used only when it is consistently accurate. A sitemap that sets priority on every URL is telling Google nothing, and a lastmod that updates when nothing changed is worse than no lastmod at all, because it teaches the crawler to distrust the field.

The measurement is in the numbers above. 6 of 89 files set changefreq, 4 set priority, and between them that is 1,574 declarations doing nothing.

03The limits that actually break a sitemap

Four hard rules, and the validator checks all of them.

50,000     maximum URL entries in one file
50 MB      maximum uncompressed size of one file
2,048      practical ceiling on the length of one loc
UTF-8      the only encoding, and loc values must be entity escaped

over any of these, split the file and list the parts in a sitemap index

The lastmod format catches more people than the size limits do. It must be W3C Datetime, which means 2026-08-24 or 2026-08-24T12:00:00+00:00. A timestamp with no timezone is invalid, and so is anything written day first. The validator checks the format exactly rather than accepting whatever parses as a date.

A sitemap index has its own rule that is easy to miss. Its entries take only loc and lastmod. Putting a changefreq or a priority inside a sitemapindex entry is invalid rather than merely ignored, and the validator reports it as an error.

04What 89 real sitemaps look like

To test the validator on real input rather than invented cases, every sitemap declared in the robots.txt of a list of major hosts was fetched on 2026-08-24 and run through the same code that runs above.

The headline is that large publishers get the mechanics right. 88 of 89 files produced no errors, and not one of them was malformed XML. What they do spend effort on is the two fields that do nothing, and a small number of duplicate entries.

One correction worth stating plainly. An earlier pass of this census reported 17 files as invalid XML. They were not. The collector stopped reading at 6 MB and cut those files off mid element, which is a fault in the measurement and not in the sitemaps. Those 17 are excluded from every figure on this page rather than counted as broken.

05The tests this validator passes

All 27 run against the code on this page, inside a real browser, because the validator depends on the browser's XML parser and testing it anywhere else would be testing something different.

07Where these rules come from

• SPECsitemaps.org protocol 0.9, for the schema, the required and optional elements, the 50,000 entry cap and the 50 MB size cap.
• SPECGoogle, build and submit a sitemap, for which fields Google reads and which it ignores.
• DATAThe census file, every figure in section 04, published unmodified.

Sitemaps and robots.txt fail together more often than either fails alone, because a sitemap that lists a path your robots.txt disallows is a contradiction Google resolves by not crawling it. The robots txt checker on this site will tell you which rule wins for any path you paste.