How to verify an email address without sending an email
Check three things in order: the address is correctly formatted, its domain publishes mail (MX) servers, and that server accepts the specific mailbox when asked. You can do all three by hand for one address; for lists, and for catch-all domains, use a verification tool or API.
By MailVet team
What verifying an email actually checks
Verifying an email address means answering one question before you send: will a message to this address reach a real mailbox? You don't need to send anything to find out. The receiving mail server gives away most of the answer in the first steps of an SMTP conversation, before any message exists.
There are three layers, each more conclusive and harder to check than the one before: format (is it shaped like an address?), domain (does the domain publish mail servers?) and mailbox (does the server accept this recipient?). The first two take seconds from any computer. The third is where the manual method gets unreliable, and where catch-all domains break it entirely.
Step 1: check the format
Most broken addresses fail here: a missing @, a space, a comma instead of a dot, a domain with no extension. Don't try to implement the full RFC 5322 grammar; it allows quoted strings and comments nobody types into a form, and strict hand-written patterns tend to reject real addresses. Code for each language is in /guides/email-validation-python, /guides/email-validation-javascript, /guides/email-validation-node and /guides/email-validation-php.
A practical format check needs only a few rules:
- One @, with something on each side of it.
- No spaces, and no dot at the start or end of the part before the @, or two dots in a row.
- A domain with at least one dot, ending in a real top-level domain such as .com, .io or .de.
- At most 64 characters before the @ and 254 in total.
- No misspelled providers: watch for gmial.com, hotmial.com, outlok.com and gmail.con.
Step 2: look up the domain's mail servers
A domain can only receive email if its DNS says where to deliver it: the MX records. Use dig on macOS or Linux, or nslookup on Windows (it works on the others too). Replace the example domain with the one you're checking.
# macOS / Linux
$ dig mx northwind.example +short
10 mx1.northwind.example.
20 mx2.northwind.example.
# Windows (Command Prompt or PowerShell)
> nslookup -type=mx northwind.example
Non-authoritative answer:
northwind.example MX preference = 10, mail exchanger = mx1.northwind.example
northwind.example MX preference = 20, mail exchanger = mx2.northwind.example
# Reading the answer
# lower number tried first; the others are backups
# no MX records senders fall back to the domain's A record (RFC 5321); rare for real mail domains
# "0 ." null MX (RFC 7505): the domain says it accepts no mail
# NXDOMAIN the domain doesn't exist, often a typoStep 3: ask the mail server about the mailbox
Connect to the first MX host on port 25 and start an ordinary SMTP conversation, but stop after RCPT TO. You type the commands; the lines that start with a number are the server's replies. You never send DATA, so no message is created and nothing reaches the inbox. The second RCPT TO uses an address that can't exist, which is how you spot a catch-all domain.
$ telnet mx1.northwind.example 25
Trying 203.0.113.25...
Connected to mx1.northwind.example.
Escape character is '^]'.
220 mx1.northwind.example ESMTP Postfix
EHLO mail.yourdomain.example
250-mx1.northwind.example
250-PIPELINING
250-SIZE 52428800
250-STARTTLS
250-ENHANCEDSTATUSCODES
250 8BITMIME
MAIL FROM:<you@yourdomain.example>
250 2.1.0 Ok
RCPT TO:<dana.reyes@northwind.example>
250 2.1.5 Ok
RCPT TO:<zq7-no-such-person-41k@northwind.example>
550 5.1.1 <zq7-no-such-person-41k@northwind.example>: Recipient address rejected: User unknown in local recipient table
QUIT
221 2.0.0 ByeWhy the manual SMTP check is unreliable, and risky
The session above is the best case: 250 for the real address, 550 5.1.1 for the made-up one, so the server checks recipients and the mailbox exists. If the made-up address had also got 250, the domain would be catch-all and the first answer would prove nothing. A 450 or 451 reply means try again later. In practice the best case is less common than you'd hope:
- Port 25 is usually blocked. Most home ISPs block outbound port 25, and cloud providers including AWS, Google Cloud and Azure block or restrict it by default. The connection just times out.
- Catch-all domains say yes to everything. Many company domains accept mail for any address, so
RCPT TOcan't tell a real person from a typo. - Accepted now, bounced later. Some servers accept every recipient during the conversation and send a bounce afterwards, once the message has been processed.
- Greylisting and rate limits. Servers often answer unfamiliar senders with a temporary 4xx code and expect a retry minutes later. Probe too often and you get throttled or disconnected.
- Your IP is judged too. Receiving servers check the connecting IP's reputation and reverse DNS. A home or unknown IP can be refused with
554or550 5.7.1before the mailbox is even looked at. - It can hurt your own mail. Lots of
RCPT TOprobes from one IP look like a directory harvest attack, the technique spammers use to discover addresses, and can get that IP blocklisted. Never run them from the server that sends your company's email. - VRFY won't help. The SMTP command designed for this is disabled on almost every public server, or answers
252: can't verify, but will try to deliver.
What each method can and can't tell you
| Method | What it tells you | What it can't tell you |
|---|---|---|
| Format check | The address is shaped like an email | Whether the domain or mailbox exists; typos like jhon@ for john@ at a real domain |
| MX lookup | The domain is set up to receive mail | Whether this mailbox exists |
| Manual SMTP check | Whether the server accepts this recipient during the handshake | Anything on catch-all domains; servers that accept and bounce later; anything at all if port 25 is blocked |
| Sending a test email | That delivery worked this once, if nothing bounces | Silence isn't proof the mail was inboxed or read, and test bounces count against your sender reputation |
| Verification API | Format, domain, mailbox and a catch-all verdict, plus disposable and role flags, each with a reason code | Future changes: a mailbox can close after it's verified. No verifier is right 100% of the time |
The practical way: a verification tool or API
A verification service runs the same three checks for you, from servers set up for the job rather than your laptop or your mail server, and handles the timeouts, retries and greylisting that make manual checks inconclusive.
MailVet also covers the step manual checks can't: on a catch-all domain it checks the specific mailbox with signals beyond the server's blanket yes, and returns deliverable with catch_all_verified or undeliverable with catch_all_no_mailbox. Anything it can't settle comes back unknown and costs nothing. No email is ever sent to the address.
- A list: upload a CSV or XLSX and download it with result columns added. See /bulk-email-verification.
- In code: one request,
GET https://api.mailvet.app/v1/verify?email=...with your API key. See /docs/quickstart. - At signup: verify in real time from your backend, covered in /use-cases/signup-form-verification.
Email verification questions
How can I check if an email address is valid without sending an email?
Check the format, look up the domain's MX records, then connect to the mail server and ask whether it accepts the recipient with RCPT TO, stopping before any message is sent. A verification tool runs the same checks for you and also handles catch-all domains, greylisting and timeouts.
Does verifying an email address notify the person?
No. The check ends before a message is created, so nothing reaches the inbox. The receiving server may log the connection, as it does for any mail traffic, but the person behind the address sees nothing.
What is a catch-all email domain?
A domain whose mail server accepts mail for any address at that domain, whether or not a mailbox exists. A manual SMTP check always gets a yes from it, so it can't tell real addresses from fake ones. MailVet checks the specific mailbox on these domains instead of stopping at 'accept-all'.
Why did an address pass an SMTP check but still bounce?
Usually one of four reasons: the domain is catch-all, the server accepts recipients first and bounces later, the mailbox was closed after you checked, or your message was refused for a reason unrelated to the address, such as authentication or sender reputation. The bounce code tells you which.
Can I verify an email address for free?
The manual steps are free if your network allows outbound port 25. MailVet gives every new account 100 free verification credits with no card, and unknown results are never charged.
Verify addresses without the telnet session
Single addresses, lists or API traffic, with catch-all addresses verified to a real answer. 100 free credits, no card.
Related
Last updated November 13, 2026