Free tool

Does this email address exist?

Checked without sending anything — and when the answer can't be known, we say so instead of guessing.

One address at a time, 5 per quarter hour from the same IP address. This limit protects the reputation of the server running the probe — ours.

How it works

  1. The way the address is written gets checked — length, characters, structure of the domain. A malformed address can't be delivered by any server: the verdict is immediate and certain.
  2. The domain must declare a mail server — we look for its MX records, and failing that its address, which acts as the server. With neither, that domain cannot receive any message at all.
  3. An SMTP probe asks whether the recipient is accepted — we announce the sender (MAIL FROM), ask for two recipients (RCPT TO): a made-up address first, yours second. Then we hang up. The DATA command, the one that delivers a message, is never sent.
  4. The verdict accounts for doubt — if the made-up address is accepted, the domain accepts everything and nothing can be concluded. We show that, rather than picking an answer at random.

Use cases

  • An address dictated over the phone or read off a piece of paper — confirm it exists before sending an important document, rather than waiting for a bounce that may never come.
  • A contact who has gone quiet — find out whether their mailbox was deleted or whether it's them keeping silent.
  • A suspicious sign-up on a form — check whether the address entered leads anywhere before creating an account.
  • Before migrating a mailbox — make sure the destination address is really accepted by its new server.
  • Diagnosing a bounce — work out whether the rejection comes from the address itself, from the domain, or from a server refusing every connection.

What this tool doesn't do

  • It never sends a message: the probe stops before delivery. The owner of the address receives nothing.
  • It can't conclude anything on a catch-all domain, which accepts every address. We detect it and say so: that's a limit of the protocol, not a fixable shortcoming.
  • It checks one address at a time, five per quarter hour from the same IP address. An SMTP probe that pushes too hard gets the server running it blocklisted: the limit protects the tool itself.
  • It doesn't tell you whether the mailbox is read, whether it's full, or whether its owner is still reachable: a server can happily accept mail for an address abandoned years ago.
  • It is not a substitute for a real send on addresses judged to exist: a spam filter or a quota can still reject an actual message later.
  • When the result is uncertain it shows "we can't tell" along with the exact reason. That's deliberate: a false certainty does more harm than no answer.

Frequently asked questions

How can you check an address without sending an email? +

By opening the beginning of an SMTP conversation with the domain's mail server: we announce a sender (MAIL FROM), ask whether the recipient is accepted (RCPT TO), and stop there. The DATA command, the one that actually transmits a message, is never sent. The recipient receives nothing and sees nothing.

What is a catch-all domain, and why does it change everything? +

A catch-all domain accepts mail for any address, whether it exists or not. On such a domain, "recipient accepted" proves nothing. We detect it by asking for a made-up address first: if that one is accepted, we answer "we can't tell" instead of selling you a false certainty.

Why is the result sometimes "we can't tell"? +

Because sometimes that's the only true answer: catch-all domain, temporary rejection (greylisting), unreachable server, session cut short, or a server that refuses our connection before the address is even mentioned. In all those cases we show the exact reason. A tool that always answers "valid" or "invalid" is simply guessing.

Can I check a whole list of addresses? +

Not with this tool: it checks one address at a time, with a limit of 5 checks per quarter hour from the same IP address. That isn't a commercial restriction but a technical one: an SMTP probe that pushes too hard gets the server running it blocklisted — ours — which would make the tool useless for everyone.

Is the address owner notified? +

No message reaches them, since nothing is delivered. The connection does show up in the logs of the domain's mail server, like any delivery attempt — unavoidable as soon as you query a server at all.

Why do some providers seem to accept everything? +

Several large providers deliberately accept every recipient at RCPT TO and sort things out after delivery, precisely to stop people enumerating their mailboxes. We detect them as catch-all and decline to conclude. It's frustrating, and it's honest.

What should I do if the verdict is "does not exist"? +

Check what you typed first: a wrong domain or two swapped letters is enough. If the address was given to you by its owner, ask them to confirm: a deleted mailbox, a former employee or an expired alias all produce the same hard rejection from the server.

Related tools