A stylesheet or script the page declares returns an error

Why it matters

Google renders pages much as a browser does and runs their JavaScript, and it notes that without rendering it might not see content that scripts bring to the page. A stylesheet or script that fails to load can change what visitors and Google see.

What to do

Look at the finding: it lists up to ten failing files with the answer each gave and the page that loads them.
Restore each file, or update the <script> or <link> to the file's current address, for example after a release gave it a new name.
If the page no longer needs the file, remove the reference; for a file on another provider's host, confirm who controls it first.

When to leave it as it is: if the failing file belongs to another provider and has no visible effect on the page, you can wait for the provider to fix it; the check does not measure a file's effect.

What we read

We read each crawled page's HTML, without running JavaScript, and collect the <script src> and <link rel="stylesheet"> it declares outside <noscript> and <template>, including files on other hosts. We then request up to 30 files per audit, starting with those most pages share; images join the same budget, and files on your own site that robots.txt forbids are not requested.

A file counts as broken when it answers 404 or 410, or a 5xx on both of our requests. Other answers, such as 401, 403, 429 or a single server error, leave it unsettled, not broken. A page is reported when at least one of its files is broken, including pages that are only a frame for scripts.

Example & evidence

Illustrative example — one page and one stylesheet, not a live audit.

Page https://bakery.example/cakes declares:

<link rel="stylesheet" href="/assets/site.4f2a.css">

GET https://bakery.example/assets/site.4f2a.css
HTTP/1.1 404 Not Found

The stylesheet answers 404, so the page is reported: 1 declared CSS or JavaScript resource returned an error. The finding shows the file, our request and its status.

A possible correction: point the href at the file the site now serves, such as /assets/site.9c1e.css, which answers 200 OK, or remove the <link> if nothing uses it.

Confirm & check again

Check for yourself: open the page with your browser's developer tools on the Network tab and look for stylesheets or scripts that fail to load.

Check again: Full audit — run a new audit of the site. The finding closes when the page is crawled again as a complete HTML page and every stylesheet and script it still declares gets a settled answer, none of them broken. It can also close when the address now answers directly with a 200 file that is not HTML, or starts with a permanent redirect within your site; problems at the new target may be reported separately. A file left unrequested by the budget, blocked by robots.txt or answering 401, 403, 429 or a single server error keeps the finding open, and so does a page that is only a frame for scripts.

Limits & sources

The check requests the files but does not run JavaScript or show how much a failing file changes the page. Files added by scripts, CSS @import, preload hints, inline scripts and fonts are not checked, and files beyond the budget of 30 are not requested.

Confidence: High because each broken status is read directly, and a server error counts only when both requests fail.

Sources

Google, How HTTP status codes affect Google's crawlers — says Google does not use content from addresses that answer 4xx and ignores content sent with a 5xx. Checked 25 September 2026.
Google, In-depth guide to how Google Search works — says Google renders pages and runs their JavaScript, and might not see content that depends on it without rendering. Checked 25 September 2026.

Editorial review: 29 September 2026

robots.txt blocks a stylesheet or script the page loads — robots.txt forbids the file, so it is not requested here.
A referenced page answers with an error — a page address, not a file, answers a client error.