501 Not Implemented: what it means and how it differs from 405
A 501 says the server doesn't recognise the request method or can't do what it asks. Learn how it differs from 405, what triggers it and how to fix it.
· 5 min read · By the Spot Downtime team
501 Not Implemented is the server saying: I don't know how to do what you asked, and that isn't specific to this URL. Most of the time that means the server doesn't recognise the request method, or it doesn't support some part of the request, such as an unusual transfer encoding.
It looks like a server error, and it is one, but it behaves differently from a 500. A 500 means something broke while handling a valid request. A 501 means the server never tried, because it doesn't have the capability.
501 vs 405
These two are easy to mix up. The difference is about scope:
- 405 Method Not Allowed: the server knows the method, but this particular resource doesn't accept it. A DELETE to a read-only route is a 405. See 405 Method Not Allowed.
- 501 Not Implemented: the server doesn't support the method at all, for any resource. It has nothing to offer you for that request, wherever you send it.
A 405 should come with an Allow header listing the methods that do work. A 501 usually has no such list, because the problem isn't the URL.
What triggers a 501
A method the server doesn't know
Web servers and frameworks recognise a fixed set of methods. Anything else can come back as 501. Typical cases are WebDAV methods such as PROPFIND or MKCOL sent to a server without WebDAV, or a custom method name sent by a misconfigured client.
Method names are case-sensitive. A client that sends get instead of GET may be answered with a 501 or a 400, depending on the server.
curl -i -X FOO https://example.com/
# Many servers answer: HTTP/1.1 501 Not ImplementedAn unsupported transfer encoding or other request feature
If a request declares a Transfer-Encoding the server doesn't understand, it can respond with 501. This shows up with hand-written HTTP clients, odd proxies or tools that build requests manually.
A proxy or gateway in front of your app
The 501 may not come from your application. A reverse proxy, load balancer or CDN can reject methods it doesn't forward. Compare a request that goes through the proxy with one that goes straight to the origin. If only the first returns 501, the fix is in the proxy configuration.
Unfinished code returning 501 on purpose
Some APIs return 501 for endpoints that are planned but not built yet. That's a legitimate use: the route exists in the design, the feature doesn't. If you see it from a third-party API, check their docs for features marked as upcoming.
Tell the layers apart with curl
Send the same request to the public address and to the origin directly, and compare the responses and the Server header. A different server name, or a different error page style, tells you which hop answered.
# Through the CDN or proxy
curl -i -X PATCH https://example.com/api/items/1
# Straight to the origin, keeping the real Host header
curl -i -X PATCH --resolve example.com:443:203.0.113.10 https://example.com/api/items/1If the origin accepts the request and the public address returns 501, the proxy or CDN is the layer that doesn't support the method. Replace the example IP with your origin's address.
How to fix it as the client
- Check the method you're sending. Look at the first column in your browser's network tab, or add
-vto curl. - Use standard uppercase methods: GET, POST, PUT, PATCH, DELETE, HEAD, OPTIONS.
- If the endpoint should accept the method, read its documentation. Some APIs use POST with a body instead of exotic methods.
- Send the same request to another URL on the same server. If every URL fails, the method itself isn't supported.
How to fix it as the site owner
Start by working out which layer answers. Run the failing request directly against the origin, then through the proxy or CDN. Check the server's access log for the request line: the log shows the exact method and protocol the server saw.
- If your app should support the method, register it in the framework or enable the module (WebDAV, for example) that handles it.
- If a proxy rejects it, allow the method in the proxy configuration, or make the client use a method that is already allowed.
- If a client library is building raw requests, update it or switch to a maintained HTTP library.
- If you return 501 deliberately for unfinished endpoints, make sure the response body explains it, and don't expose these routes in public docs until they work.
501 is not transient
Monitoring for it
A 501 on a page you monitor is rare and almost always means a configuration change: a new proxy rule, a server upgrade or a client sending a different method. A monitor that expects 200 will flag it immediately, and the response in the alert tells you which layer answered. If you check an API endpoint that needs a specific method, set that method on the monitor so it exercises the real route. Spot Downtime's API monitoring lets you choose the method, headers and body.
Get new guides by email
Practical posts on monitoring, outages, error fixes and on-call, when they're published. No spam.
Keep reading
- HTTP errors · Troubleshootingnginx 502 Bad Gateway: fixing "connect() failed" to PHP-FPM and other upstreamsnginx logs connect() failed (111), (2) or (13) when it can't reach your app. What each errno means, and how to fix PHP-FPM sockets, ports, permissions and SELinux.November 13, 2026 · 7 min read
- HTTP errors · Troubleshooting407 Proxy Authentication Required: what it means and how to fix itA 407 comes from a proxy, not the website. Why it appears in browsers, curl and CI jobs, how it differs from 401, and how to send proxy credentials correctly.November 12, 2026 · 6 min read