How to check if a website is down? The guides all cover the same tools, but none of them put the steps in the one order that actually diagnoses the problem fastest. That order is: rule out your side in thirty seconds, get an external verdict, then read the error code. Follow it in that sequence and you will know within two minutes whether the site is down, your network is at fault, or your browser is lying to you.

Start with the thirty-second checks. Refresh the page and wait a minute, because temporary server hiccups pass on their own. Try another website: if google.com loads fine, your internet connection works and the problem is specific to the site you want. Switch networks, Wi-Fi to mobile data or vice versa: if the site loads on one network but not the other, the problem sits between your network and the server, not on the server itself. Try a different browser or an incognito window, because extensions, corrupted caches and stale cookies can block a page that is perfectly healthy.

Next, get an external verdict from a checker that tests from its own servers, not your connection. Downforeveryoneorjustme.com gives a simple up or down answer. Isitdownrightnow.com adds response time and recent history. DownDetector adds crowd-sourced outage reports, graphs and heat maps, which catch partial outages a single server check can miss. Site24x7 checks from multiple global locations, which matters because a site can be down in one region and fine in another. If the external checker says the site is up but your browser still fails, the problem is on your end: DNS, your network, a VPN, or a firewall.

Then read the error code, because it names the failure. A 200 means the site is up. A 404 means the page is missing but the server works. A 500 means an application error on the server. A 502 or 504 means a backend or gateway failure, often a crashed app server. A 503 means the server is overloaded or in maintenance. A 403 can mean access blocked, but note that some sites answer automated checks with 403 or 405 even when they work fine for people, so treat those as up unless a browser also fails. A timeout or a DNS error means the server is not answering at all: check nameservers, domain expiry and DNS records.

For a deeper check, use the command line. Ping the domain: replies mean the server is online, timeouts mean it is not answering, though some major sites refuse pings entirely, so a failed ping alone proves nothing. Traceroute shows where the connection fails on its path. Curl returns the raw HTTP status code without browser interference. And your browser's developer tools, opened with F12, show the exact failing request on the Network tab, which is what support teams actually need from you.

What to do depends on which side the fault is on. If you are a visitor and the site is genuinely down, there is nothing to fix on your machine: wait, check DownDetector for outage reports, and try again later. If it is your side, flush your DNS cache, switch DNS servers, disable your VPN, or try another network. If you own the site, a 5xx code means checking your app server logs and restarting a crashed backend, a DNS failure means checking nameservers and expiry, and a regional outage means checking your hosting provider and CDN origin health. Set up uptime monitoring afterwards, so the next outage alerts you instead of your visitors telling you.

Bottom line: thirty-second local checks, one external checker, then the error code. In that order, you will know in two minutes whether a website is down for everyone or just for you, and exactly what to do about it.