ERR_CONNECTION_TIMED_OUT: Why It Happens and How to Fix It

ERR_CONNECTION_TIMED_OUT shows up when your browser sends a request and simply never gets an answer back — not an error response, not a redirect, nothing. After roughly 30 seconds of silence, it gives up and shows you this page. That silence is the key clue: unlike a DNS error or an SSL warning, the connection never got the chance to properly start.

What’s different about a timeout

Most connection errors happen because something answered and said “no” — a bad certificate, a missing DNS record, a 404. A timeout is the opposite: nothing answered at all. That puts the likely causes in a different place than usual — somewhere along the network path, or on an overloaded server, rather than in a specific configuration mistake.

Step 1: Figure out the scope

Before changing anything, narrow down where the silence is happening:

• Try other sites. If nothing loads, the problem is your network or device, not the one site.
• Try the same site on mobile data, Wi-Fi off. This swaps your entire network path. If it loads instantly on mobile data, your original network — router, ISP, or a captive portal — is the bottleneck.
• Run it through a server-side checker. This sends the request from an entirely different network with no relationship to yours. If the independent check succeeds, the timeout is local to you. If it also times out, the server itself is the problem.

Step 2: If it’s local to you

Restart your router and modem. Power them off, wait about 30 seconds, and power them back on. This clears a surprising number of routing and DHCP glitches that produce exactly this symptom.

Check for a VPN or proxy. These add a hop that can itself be slow or unreachable. Disable temporarily and retry.

Disable browser extensions, especially ad blockers or privacy extensions, one at a time. Some aggressively intercept requests in ways that produce a timeout instead of a clean block.

Clear your browser cache. A corrupted cache entry for a site occasionally causes requests to hang rather than fail cleanly.

Check your firewall or antivirus. Security software sometimes holds connections open while it inspects them, long enough to trigger a timeout on slower connections.

Step 3: If it’s the server’s side

If an independent check also times out, the server isn’t answering for anyone, not just you. Common causes, roughly from most to least common:

• The server is overloaded. A traffic spike, a resource-heavy process, or simply too many simultaneous requests for the hosting plan’s resources — this is especially common on shared hosting, where one busy neighbor can slow everyone down.
• A firewall or security rule is silently dropping the connection. Unlike a proper rejection, a dropped packet produces exactly this “no response at all” symptom instead of a clean error.
• The server process has crashed or hung, and nothing is listening on the port anymore.
• A misconfigured DNS record is pointing the domain at a server or IP that no longer exists or no longer answers on that port.

If it’s your own server: check your host’s resource graphs for CPU/memory spikes around the time of the outage, check that the web server process (nginx, Apache, etc.) is actually running, and review firewall/security group rules for anything that might be dropping inbound connections.

The quick diagnostic

The fastest way through this error is the same one-question test every connection issue reduces to: does an independent server reach it? A live status check answers that in a couple of seconds — reachable there but not for you means it’s local (Step 2); unreachable there too means the server needs attention (Step 3). Skipping that question is how people end up restarting their router for a problem that was never theirs, or waiting on a site that was fine all along.

For the DNS-flavored version of “can’t connect,” see DNS_PROBE_FINISHED_NXDOMAIN; for the certificate-flavored version, see “Your connection is not private”.