Skip to content

Google News sitemap: required tags and the two-day rule

How to build a Google News sitemap: the required news tags, the rule to list only articles from the last two days, the 1,000-entry limit and the errors to fix.

By Crawlito · 8 min read

In short

  • A Google News sitemap is a normal XML sitemap with the news namespace, where each article URL carries one news:news block holding the publication name, language, publication date and title.
  • Google says to list only articles created in the last two days, and after that to remove the URLs or remove their news:news metadata.
  • One sitemap file can hold at most 1,000 news:news tags; Search Console reports more than that as the Too many News URLs error.
  • news:publication_date is the date the article was first published, not the date it was added to the sitemap, in a W3C format such as 2026-11-15T08:30:00+00:00.
  • Google says to add new articles to the existing news sitemap as they are published rather than creating a new sitemap with each update.
On this page

A Google News sitemap is an ordinary XML sitemap that adds a news:news block to each article URL, with four facts Google requires: the publication name, the language, the first publication date and the article title. Google's news sitemap documentation adds two rules that make it different from every other sitemap you run: list only articles created in the last two days, and keep no more than 1,000 news entries in one file. Update the same file as you publish instead of generating a new one each time.

This post covers only what is specific to news sitemaps. For the general protocol, size limits and submission, see XML sitemaps for AI discovery. Whether Google News shows your publication at all is a separate question, covered in submitting a site to Google News: a news sitemap tells Google about recent articles; it does not make a site eligible or guarantee placement.

What a valid Google News sitemap looks like

Here is a complete file with two articles. It follows the structure of Google's own example, with an https URL and an escaped ampersand in the second title.

xml
<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"
        xmlns:news="http://www.google.com/schemas/sitemap-news/0.9">
  <url>
    <loc>https://www.example.com/business/2026/11/15/port-strike-ends</loc>
    <news:news>
      <news:publication>
        <news:name>The Example Times</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-15T08:30:00+00:00</news:publication_date>
      <news:title>Port strike ends after nine days of talks</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://www.example.com/tech/2026/11/14/chipmaker-results</loc>
    <news:news>
      <news:publication>
        <news:name>The Example Times</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-14T17:05:00+00:00</news:publication_date>
      <news:title>Chipmaker beats forecasts as R&amp;D spending rises</news:title>
    </news:news>
  </url>
</urlset>

Two details catch people out. The news prefix only works because the namespace is declared on urlset, and the value must be exactly http://www.google.com/schemas/sitemap-news/0.9. Search Console reports a missing or misspelt namespace on the root element as Incorrect namespace, and an undeclared news prefix makes the file invalid XML. And because the sitemaps protocol requires UTF-8 and entity escaping for &, ', ", < and > in every data value, a headline with a raw ampersand breaks the whole file, not just one entry.

The required news tags, one by one

Every tag below is required by Google. Each <url> may contain only one news:news block, and each news:news only one news:publication.

TagWhat goes in itRules Google documents
news:newsParent of the other news tagsOne per <url>
news:publicationParent of news:name and news:languageOne per news:news
news:nameYour publication nameMust match the name as it appears on news.google.com, without any text in parentheses
news:languageArticle languageISO 639 code of two or three letters; the exceptions are zh-cn (Simplified Chinese) and zh-tw (Traditional Chinese)
news:publication_dateWhen the article was first publishedW3C format: YYYY-MM-DD, YYYY-MM-DDThh:mmTZD, YYYY-MM-DDThh:mm:ssTZD or with fractional seconds
news:titleThe headlineAs it appears on your site, without author name, publication name or date; Google may shorten it on some devices

Some practical notes on the fields that cause the most trouble:

  • news:name. Use one fixed string from configuration, not a value assembled from the page template, so it cannot drift between "The Example Times", "Example Times" and "The Example Times (UK)". Google's Publisher Center help also says Google News will use site names for publication titles, so keep the sitemap value consistent with the site name your homepage declares.
  • news:language. Use the language of the article, not of the site. A bilingual publication can mark one entry en and the next es in the same file. Regional variants such as en-gb are not in the documented list; only the two Chinese codes carry a region.
  • news:publication_date. This is the first publication time. Google is explicit that it is not the date the article was added to the sitemap. A time zone designator (Z or +01:00) matters: a timestamp with a time but no zone is not one of the documented formats, so do not rely on Google accepting it. Keep it consistent with the date in your Article structured data and the visible byline.
  • news:title. Take the headline as shown on the page, not the SEO title with " | The Example Times" appended.

The two-day rule and how to apply it

Google's guidance is short: include only articles created in the last two days, and once an article is older, either remove its URL from the news sitemap or remove its news:news metadata. Google does not say whether "two days" means calendar days or 48 hours. A rolling 48-hour window based on news:publication_date satisfies either reading.

That rule changes how the file behaves compared with a normal sitemap:

Normal XML sitemapGoogle News sitemap
What it listsEvery canonical URL you want crawledOnly articles created in the last two days
Per-file limit50,000 URLs, 50 MB uncompressed1,000 news:news tags (and the general limits)
Date that matterslastmod, the last significant changenews:publication_date, the first publication
Over timeGrows with the siteRolls: entries enter and leave
Can be emptyRarely, and usually a bugNormal on quiet days

If your newsroom goes quiet for two days, the file may legitimately contain no entries. Google says Search Console may then show an Empty sitemap warning, and that an empty news sitemap does not harm your site in Google Search. Keep the file served with a 200 status and a valid empty urlset, rather than deleting it or returning a 404.

Older articles do not disappear from Google because they leave the news sitemap. They belong in your regular sitemap, where lastmod does the work described in keeping sitemap lastmod honest. You can also put the news tags into your existing sitemap instead of a separate file: Google allows both, and says a separate news sitemap can make tracking in Search Console easier. If you combine them, strip the news:news block from entries older than two days while keeping the URL.

Generate it from your article records

A news sitemap should be built from the same publication records as your pages and feed, filtered by first publication time and capped at 1,000 entries. This sketch in TypeScript shows the logic; adapt the data source to your CMS.

typescript
type Article = { url: string; title: string; lang: string; firstPublished: Date };

const PUBLICATION = "The Example Times"; // must match news.google.com
const WINDOW_MS = 48 * 60 * 60 * 1000;
const MAX_ENTRIES = 1000;

const esc = (s: string) =>
  s.replace(/&/g, "&amp;").replace(/</g, "&lt;").replace(/>/g, "&gt;")
   .replace(/"/g, "&quot;").replace(/'/g, "&apos;");

export function newsSitemap(articles: Article[], now = new Date()): string {
  const recent = articles
    .filter((a) => now.getTime() - a.firstPublished.getTime() <= WINDOW_MS)
    .sort((a, b) => b.firstPublished.getTime() - a.firstPublished.getTime())
    .slice(0, MAX_ENTRIES);

  const entries = recent.map((a) => `  <url>
    <loc>${esc(a.url)}</loc>
    <news:news>
      <news:publication>
        <news:name>${esc(PUBLICATION)}</news:name>
        <news:language>${a.lang}</news:language>
      </news:publication>
      <news:publication_date>${a.firstPublished.toISOString()}</news:publication_date>
      <news:title>${esc(a.title)}</news:title>
    </news:news>
  </url>`);

  return `<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"
        xmlns:news="http://www.google.com/schemas/sitemap-news/0.9">
${entries.join("\n")}
</urlset>`;
}

Three choices in that code follow Google's rules directly. The filter uses the first publication time, so editing an old story does not pull it back into the window. The cap keeps one file under 1,000 entries; if you publish more than that in two days, split the output into several news sitemaps rather than dropping articles, as Google advises. And the file is regenerated at one fixed URL, because Google says to add fresh articles to the existing news sitemap and not to create a new sitemap with each update. Google also says Google News crawls news sitemaps as often as it crawls the rest of your site, so there is no separate schedule to request.

Serve the file at a URL inside the scope of the articles it lists. The sitemaps protocol says a sitemap can list only URLs on the same host and protocol, under its own path, so /news-sitemap.xml at the root works for the whole site, while /sport/news-sitemap.xml can list only /sport/ URLs. Google's sitemap build guide makes one exception: a sitemap submitted through Search Console is not limited to its own directory. The root is still the simplest choice. Then add it to robots.txt with a Sitemap line or submit it in Search Console like any other sitemap.

News sitemap or RSS feed

Google accepts RSS 2.0 and Atom 1.0 feeds as sitemaps and notes that a feed only provides information on recent URLs, which sounds close to a news sitemap. The difference is the metadata. A feed gives a URL, a title and a date, but no news:name, no news:language and no defined rule that the date is first publication. The news sitemap is the format Google documents for telling it about news articles specifically.

Feeds have also lost a role they used to have. Google's Publisher Center help says that from late March 2025, Google News no longer uses RSS feeds or web locations submitted in Publisher Center. Keep your feed for readers, aggregators and other crawlers (see RSS feeds for content discovery), and use the news sitemap for Google.

News-specific errors in the Search Console Sitemaps report

Google's sitemap errors reference lists two errors that only apply to news sitemaps, plus general ones that news files trigger often.

ErrorCauseFix
Too many News URLsMore than 1,000 news URLs in one fileApply the two-day filter, then split what remains across files
Missing <publication> tagA URL in the news sitemap has no news:publicationEmit news:publication with name and language for every entry
Invalid dateA date that is malformed or not a real calendar dateOutput one of the W3C Datetime formats, with a time zone designator when you include a time
Parsing errorXML that cannot be parsed, often an unescaped & in a headlineEscape every data value
Incorrect namespaceThe namespace on the root urlset element is missing, misdeclared or misspeltCopy both namespace URLs exactly from Google's example
Empty sitemapNo articles in the last two daysNothing, if your publishing really paused

If the report says the file itself could not be retrieved, the cause is outside the XML; work through sitemap couldn't fetch in Search Console.

A quick check from the command line catches most of the table before Google does:

bash
curl -s https://www.example.com/news-sitemap.xml -o news.xml
xmllint --noout news.xml && echo "well-formed"
grep -c "<news:news>" news.xml                      # must be 1000 or fewer
grep -o "<news:publication_date>[^<]*" news.xml | cut -d'>' -f2 | sort | head -3

The last line prints the oldest dates in the file, as long as every date uses the same time zone form (the generator above always writes UTC with Z). If any is more than two days old, your filter is using the wrong date field.

Frequently asked questions

Is a Google News sitemap required to appear in Google News?

Google's news sitemap page does not describe it as mandatory; it says publishers can use one to tell Google about their news articles. Google's Publisher Center help says content that follows its content policies is automatically eligible for consideration, so the sitemap does not grant eligibility.

Can I add news tags to my existing XML sitemap?

Yes. Google allows news tags in an existing sitemap or a separate news-only sitemap, and says a separate file can make tracking in Search Console easier.

What happens to articles after two days?

Remove them from the news sitemap, or remove their news:news block if you use a combined sitemap. Keep them in your regular sitemap so they stay discoverable as normal pages.

Should news:publication_date change when I update an article?

No. Google defines it as the date the article was first published. Record significant updates in lastmod in your regular sitemap and in your article's modified date instead.

Do AI crawlers read Google News sitemaps?

The news extension is documented by Google for Google. Other crawlers that read sitemaps can still pick up the loc URLs, but no AI crawler operator documents using the news tags.

Check the articles behind the sitemap

A clean news sitemap only helps if the articles it lists return 200, are not blocked and show their content in the HTML. Run the free check on a recent article URL to see whether Googlebot and the main AI crawlers can fetch and read it.

Filed under Structure and schema. Topics: sitemaps, google, content-freshness