Skip to content
MailVet
Glossary

How does SMTP email verification work?

SMTP email verification checks whether a mailbox exists by connecting to the domain's mail server, starting a normal delivery conversation, and asking whether it would accept mail for the address. It disconnects before any message is sent, so nothing reaches the inbox.

Every email delivery starts with an SMTP conversation: the sender introduces itself, names the envelope sender, names the recipient, and only then sends the message. The receiving server replies to each step with a code. SMTP verification runs the first part of that conversation, stops at the recipient step, and reads the server's answer as a verdict on the mailbox.

It's the core of every email verifier, and it's only as good as the server's answer: some servers accept every recipient, some defer strangers, and some limit how many questions they'll answer.

How it works

The checks, in order

  1. 01

    Syntax

    Is the address well-formed? Malformed addresses stop here, before any network check. See /glossary/email-syntax.

  2. 02

    DNS and MX

    Does the domain exist, and which servers receive its mail? See /glossary/mx-record.

  3. 03

    SMTP handshake

    Connect to the preferred MX host on port 25, then send EHLO, MAIL FROM, and RCPT TO with the address being checked.

  4. 04

    Read the reply, then quit

    250 means accepted, 550 means no such user, 4xx means try later. QUIT comes before DATA, so no message is ever sent.

A verification conversation

C is the verifier, S is the receiving server. It ends before DATA, the command that would start an actual message.

SMTP session (abridged)
S: 220 mx1.northwind.example ESMTP ready
C: EHLO verify.example
S: 250 mx1.northwind.example
C: MAIL FROM:<probe@verify.example>
S: 250 2.1.0 Sender OK
C: RCPT TO:<dana@northwind.example>
S: 250 2.1.5 Recipient OK        <- mailbox accepted
C: RCPT TO:<dnaa@northwind.example>
S: 550 5.1.1 User unknown        <- no such mailbox
C: QUIT
S: 221 2.0.0 Bye

What the reply codes mean for a mailbox

ReplyMeaningMailVet result
250Recipient accepteddeliverable · accepted, unless the domain turns out to be catch-all
550 5.1.1, 551No such user, or user not on this serverundeliverable · mailbox_not_found
452 4.2.2, 552Mailbox over quotarisky · mailbox_full
553Mailbox name not allowedUsually caught earlier as invalid_syntax
421, 450, 451Temporary: greylisting, rate limiting, server busyunknown · no_response, free
No replyTimeout or connection refusedunknown · no_response, free

A 550 with a 5.7.x enhanced code is a policy block on the connecting server, not a verdict on the mailbox.

Where a plain SMTP check falls short

Some gaps can be handled with more work, and MailVet handles them. Others no SMTP check can close.

  • ✓Catch-all domains accept every RCPT TO. MailVet runs a deeper check and returns catch_all_verified or catch_all_no_mailbox.
  • ✓Greylisting defers the first attempt. MailVet returns unknown, free, rather than guessing.
  • ✓Slow or silent servers time out. Also unknown, also free, and worth re-running later.
  • ✓Disposable domains may accept mail today and vanish tomorrow. They come back risky with reason disposable.
  • ×No check can tell whether a person actually reads the mailbox
  • ×An accepted address can still be a spam trap, because traps accept mail like any inbox
  • ×A check is a snapshot: a mailbox that exists today can be closed next month
FAQ

SMTP verification questions

Does SMTP verification send an email?

No. The conversation stops before the DATA command, which is where a message would begin. The recipient's inbox never receives anything, and the person isn't notified.

Is SMTP email verification accurate?

On servers that answer honestly, it's a direct answer from the server that would receive the mail. Accuracy drops on catch-all domains, greylisting servers and servers that limit checks, which is why good verifiers report those cases separately instead of calling them valid. No verifier is 100% accurate.

Can I do SMTP verification myself?

You can test a single address by hand over port 25, but many networks block outbound port 25, and mail servers often throttle or block IPs that make repeated checks. For whole lists, a verification service is more reliable and keeps your own server's reputation out of it.

Why do some verifications come back unknown?

Because the server didn't give a usable answer: it timed out, greylisted the check, or sat behind a catch-all setup where the evidence didn't agree. MailVet returns unknown in those cases and doesn't charge for it.

Skip the port 25 plumbing

One API call runs every check above and returns a result, a reason and the MX host. 100 free credits to start.