Every time you type a web address and hit Enter, something invisible happens in the background in a fraction of a second — a DNS lookup. It's one of the most important processes on the internet, and also one of the least understood. Yet once you get how it works, a whole category of everyday problems ("this site won't load," "DNS error," "why is my browsing slow?") suddenly makes sense. Here's a plain-English guide to what a DNS lookup is and how it works, no networking degree required.

What is a DNS lookup?

Computers don't actually find each other by name — they find each other by IP address, a string of numbers like 93.184.216.34. But nobody wants to memorize numbers for every website, so we use friendly names like example.com instead. A DNS lookup is the process of translating that human-friendly name into the numeric address your device needs to connect.

Think of DNS — the Domain Name System — as the internet's giant, distributed phonebook. You know the name of who you want to call, but you need their number. A DNS lookup is the act of flipping through that phonebook to find it. The clever part is that this "phonebook" isn't one single book — it's a chain of servers around the world that hand your question along until someone answers it.

The players involved in a DNS lookup

A handful of participants work together during every lookup:

  • Your browser and device cache — the first place checked, holding recently used addresses.
  • The recursive resolver — usually run by your ISP (or a public one like 1.1.1.1), this is the "middleman" that does the legwork of finding the answer for you.
  • Root nameservers — the top of the hierarchy, which point toward the right directory for each domain ending.
  • TLD nameservers — responsible for a specific top-level domain like .com, .org, or .net.
  • Authoritative nameservers — the servers that hold the actual, official address for the specific domain you asked about.

How a DNS lookup works, step by step

Here's the journey your request takes when you visit a site for the first time:

  1. You enter the address. Your browser first checks its own cache to see if it already knows the answer. If it does, you're done instantly.
  2. The question goes to the recursive resolver. If your device doesn't know the address, it asks the recursive resolver to find it.
  3. The resolver asks a root nameserver. The root server doesn't know the exact address, but it knows who handles the domain's ending — for example.com, that's the .com directory.
  4. The resolver asks the TLD nameserver. The .com server points the resolver to the authoritative nameserver for example.com.
  5. The resolver asks the authoritative nameserver. This server holds the real answer and returns the IP address.
  6. The resolver hands the address back to your device, which finally connects to the website's server and loads the page.

All of this typically happens in well under a tenth of a second.

Caching: why the second visit is faster

You don't repeat that whole journey every time. Along the way, each participant caches (stores) the answer for a set period, defined by a value called the TTL (Time To Live). The next time you — or anyone using the same resolver — visits the site, the cached answer is served instantly, skipping the full lookup.

This is why the first visit to a new site can feel a touch slower than later ones. It's also why, when a website changes servers, you might reach the old address for a while: your cache is still holding the previous answer until its TTL expires. Clearing that cache forces a fresh lookup — our guide on how to flush your DNS cache explains exactly how.

The records a lookup can return

A DNS lookup can ask for different record types, depending on what's needed:

  • A record — maps a name to an IPv4 address (the classic lookup).
  • AAAA record — maps a name to an IPv6 address.
  • CNAME — points one name to another name (an alias).
  • MX — tells email where to go for that domain.
  • NS — lists the authoritative nameservers for the domain.

You can see all of these for any domain using our free DNS Records tool — just enter a domain and it queries the authoritative servers live, showing you exactly what a real DNS lookup returns.

When a DNS lookup goes wrong

Because so many steps are involved, a DNS lookup can fail in several ways — and each produces a recognizable error. If the name simply can't be resolved, you'll see errors like ERRNAMENOTRESOLVED or DNSPROBEFINISHEDNXDOMAIN, which usually mean a stale cache, an unreachable resolver, or a domain that doesn't exist. Often the fix is as simple as flushing your cache or switching to a faster public resolver like 1.1.1.1.

If a site won't load and you suspect DNS, you can also run it through our live website down checker to confirm whether the domain resolves from our servers — telling you instantly whether the problem is the domain itself or just your local setup.

The bottom line

A DNS lookup is simply the internet looking up a name to find its number — a rapid relay from your device, to a resolver, to the root, TLD, and authoritative servers, and back again, usually in milliseconds. Caching keeps repeat visits fast, TTLs control how long answers stick around, and when something in the chain breaks, you get one of the familiar DNS errors. Understanding this one invisible process makes a huge share of everyday connection problems far easier to diagnose — and far less mysterious.