How greylisting works

Greylisting is described in RFC 6647. When a message arrives, the receiving server builds a record from three values, often called the triplet:

  • the connecting client’s IP address,
  • the envelope sender (MAIL FROM),
  • the recipient (RCPT TO).

If the server has not seen that combination before, it does not accept or reject the message. It returns a temporary failure, a 4xx reply as defined in RFC 5321, and stores the triplet with a timestamp. A typical exchange looks like this:

C: MAIL FROM:<news@sender.example>
S: 250 2.1.0 OK
C: RCPT TO:<jane@example.com>
S: 450 4.7.1 Greylisted, please try again later

The exact code and text vary by server; 450 and 451 with a 4.7.x or 4.2.0 enhanced code are common. A properly configured sending server queues the message and retries. RFC 5321 suggests retry intervals of at least 30 minutes, although many senders retry sooner. When the retry arrives within the receiver’s accepted window, the triplet passes and the message is delivered. RFC 6647 recommends that, by default, a retry be accepted if it arrives between one minute and 24 hours after the first attempt.

Many implementations then remember the sender. RFC 6647 suggests that after a successful retry, further traffic from that IP address can be allowed without delay, so greylisting usually affects only the first message from a new source. Some servers track whole IP ranges instead of single addresses, which helps when a sender delivers from a pool of servers.

Why greylisting matters for senders

For senders with correctly configured mail servers, greylisting mostly causes a short delay on first contact. It becomes a real problem when:

  • Your sending software does not retry, or gives up too quickly. Messages are then lost, and the failures look like soft bounces (see what is a soft bounce).
  • You send from a large, rotating pool of IPs. Each retry may come from a different address, so the triplet never matches and the delay repeats.
  • Your messages are time sensitive. Password resets and one-time codes can arrive several minutes late on greylisting domains.

Greylisting is not a reputation penalty in itself, but servers that combine it with reputation data may greylist unknown or poorly rated senders more aggressively. Keeping a stable sending setup and a healthy sender reputation helps.

Why greylisting matters for email verification

Verification relies on the server’s response to RCPT TO (see how SMTP verification works). A greylisting server answers with a 4xx deferral, so a verifier that tries only once learns nothing about the mailbox. Poor tools then make one of two mistakes: they report the address as invalid, which leads you to delete real contacts, or they report it as unknown when a retry would have produced a clear answer.

How to handle greylisting

As a sender:

  1. Use a mail server or email service that queues and retries temporary failures automatically.
  2. Send each message’s retries from the same IP address, or a small consistent set, so the triplet matches.
  3. Treat 4xx replies as deferrals, not hard bounces, and only suppress an address after it fails permanently.

As a verifier or list owner:

  1. Never treat a greylisting reply as undeliverable.
  2. Retry greylisted checks after a delay, from the same IP, before settling on a result.
  3. If a verdict is still unavailable, keep the address as unknown and re-check it later rather than deleting it.

Greylisting vs. rate limiting

The two are often confused. Greylisting is a one-time “prove you will retry” test tied to a specific sender, recipient, and IP combination. Rate limiting caps how many connections or messages an IP can send in a time window, regardless of retry behavior. Both produce 4xx replies, and both require patient, well-paced retries.

Where RocketVerifier helps

When a receiving server defers a check, RocketVerifier reports the address as unknown with a greylisted sub_status rather than undeliverable, so real contacts are not deleted by mistake, and unknown results are never charged. Explainable verification shows the reason behind each result, and the bounce code reference explains the deferral codes you may see in your own sending logs.