How we count
We find the sitemap (through robots.txt, or /sitemap.xml), follow every index into its child sitemaps, decompress gzip files, and count the <url> entries in each file. Then we count unique URLs across all files. Two rules keep the total honest:
- Each file is read once. robots.txt often lists a sitemap index and its children. A file reached twice (from robots.txt and from an index, or through a redirect to a file we already read) isn’t counted again.
- Each URL is counted once. A URL in two files counts toward the first file’s “new URLs” and shows as a duplicate in the second, so the “new URLs” column adds up to the unique total.
Image and video URLs inside sitemap extensions are not counted: only each entry’s page URL.
Why the number is worth knowing
- Index coverage. Compare it with the indexed count in Search Console. If you submit 5,000 URLs and 1,200 are indexed, look at the “Crawled, currently not indexed” reasons before publishing more.
- Growth tracking. Programmatic and content sites grow in batches. Counting per file shows which section grew.
- Sanity checks. A sitemap with far more URLs than the site has pages often includes parameters, drafts or duplicates; one with far fewer is missing sections.
- Limits. A single file may hold 50,000 URLs and 50 MB uncompressed. Past that, split it and list the parts in an index.
To see what those URLs are, use the URL extractor; to see how they break down by section and date, the sitemap analytics tool.