Check redirects and HTTP status codes

How to use the tool

Paste one or more URLs, one per line, then click “Check”. For each URL, the tool sends a request, records the server’s response and, when it gets a redirect, starts again with the destination given in the Location header. It stops at the first response that is not a redirect, or after ten redirects.

Each URL gets a table with every hop: the URL requested, the status code returned, the destination and the response time measured from the tool’s server. The final URL sits under the table and the alerts above it. The “Download as CSV” button saves everything in a file that opens in any spreadsheet.

URLs are checked one after the other. The tools on this site accept twenty requests per ten-minute window and per visitor, all tools combined: past that, the check stops and lists the URLs left aside.

The tool’s alerts

Each alert rests on a rule published by Google or on the HTTP standard.

• Loop. A URL comes back in its own chain. The HTTP standard asks clients to detect these loops and intervene⁵: the browser shows an error and the crawler never reaches a page.
• More than ten redirects. Google’s crawlers follow up to ten redirect hops.¹ The tool stops at the same point.
• Redirect chain. From two redirects in a row, the tool flags it: Google advises redirecting straight to the final destination, and otherwise keeping chains short, ideally no more than three redirects and fewer than five.³
• Temporary redirect. A 302, 303 or 307 in the chain. If the move is permanent, a 301 or 308 sends Google a clearer signal.²
• HTTPS to HTTP. Google prefers the HTTPS version of a page as canonical, except, among other cases, when it redirects users to or through an HTTP page.⁴
• Destination in error. The chain ends on a 4xx or 5xx response: the redirect sends visitors and crawlers to a page that is missing or down.
• Redirect to the home page. An inner page that ends up on the home page may be treated as a soft 404.³

Permanent or temporary redirect

The HTTP standard, RFC 9110, sorts redirects into two families. A 301 or 308 says the resource has been given a new permanent URI; a 302 or 307 that it temporarily lives at another URI.⁵ A 303 points to another resource, usually after a form has been submitted.

The difference between 301 and 308, as between 302 and 307, lies in the request method. For historical reasons, a browser may turn a POST into a GET when it follows a 301 or a 302⁵; 307 and 308 were added to rule that out.⁶ For a page visited the normal way, with GET, 301 and 308 are interchangeable.

On Google’s side, permanent codes are a strong signal that the target should become the canonical URL, temporary codes a weak one.¹ Google recommends a permanent redirect when you are sure it will not be reverted, and a temporary one to keep the old URL in results, during an outage for instance.² A 302 set by mistake on a permanent move therefore leaves Google free to keep the old URL.

How many redirects Google follows

By default, Google’s crawlers follow up to ten hops. Any content received from a redirecting URL is ignored; only the final destination’s content is processed.¹ A chain of four or five redirects is still followed by Googlebot, but Google advises against it: point straight to the final destination, or else keep the chain short, because each hop slows visitors down and not every browser follows long chains.³

Chains usually build up over time: a redirect from HTTP to HTTPS, then another to the www version, then a third added when a URL changes. The address nemociment.fr/blog requested over HTTP is a simple example: a 301 to HTTPS, then a second one to /blog/ with the trailing slash. The fix is to point each old URL straight to the last one, and to update internal links so they target the final URL.

When the chain ends on an error

A redirect is only useful if it lands on a page that returns 200. Google handles all 4xx errors the same way, except 429: the page is treated as missing, removed from the index if it was there, and crawled less and less often. 5xx errors and 429 slow crawling down; pages already indexed stay for a while, then are dropped if the error persists.¹

A page deleted with no equivalent can return 404 or 410. A page that has moved should redirect to its new URL, and that new URL should return 200.

Redirecting to the home page

During a redesign, it is tempting to send every old page to the home page. Google advises against it: redirecting many old URLs to one irrelevant destination, such as the home page of the new site, confuses users and might be treated as a soft 404.³ The tool flags it as soon as an inner URL ends up at the root of a site. The right target is the page closest to the old one; failing that, a real 404.

Google also advises keeping these redirects for as long as possible, generally at least one year, so that signals can move to the new URLs.³ That is the core of a website redesign that keeps its rankings: a page-by-page redirect plan, checked URL by URL before and after launch.

What the tool does not see

The tool reads server-side redirects, with a 3xx code and a Location header. Two other methods exist. A meta refresh tag sits in the HTML: Google treats an instant one as permanent and a delayed one as temporary. JavaScript redirects are only seen when the page is rendered, after crawling, and Google advises using them only when a server-side redirect is not possible.² Browsers apply the HTTP redirect first, then the meta tag, then JavaScript.⁶

A page that returns 200 can also be blocked in other ways, by a noindex tag or by robots.txt. For the latter, the AI crawler access test reads a site’s file and tells you, crawler by crawler, whether a page can be read. Finally, the tool queries sites from its own server: a site that serves a different response by country or browser may answer your visitors differently.

Frequently asked questions