Delivery Status Notification Failure Error: Causes and Fixes

A delivery status notification failure error lands in your inbox when a message you sent could not be delivered. Gmail labels these bounce-backs with the subject line "Delivery Status Notification (Failure)", while Microsoft 365 and Exchange call them non-delivery reports (NDRs). Under the hood, both are the same standardized message type, and both carry the exact reason your email failed — if you know where to look.

This guide explains what these notifications actually are, how to tell a genuine bounce from a forged one, and how to fix each common cause: bad addresses, spam and policy rejections, sending limits, full mailboxes, mail loops, and authentication failures. Whether you are troubleshooting a single Outlook message or handling thousands of bounces in a transactional mail pipeline, reading the error correctly is the fastest path to a fix.

Table of Contents#

What a Delivery Status Notification Really Is#

A delivery status notification (DSN) is an automated report generated by a mail server when it cannot deliver, relay, or gateway a message. The most familiar form is the non-delivery report: a "failed" DSN telling you the message was abandoned. Servers can also send "delayed" DSNs while they keep retrying, and — if the sender explicitly requested it — "delivered", "relayed", or "expanded" confirmations.

The format is defined in RFC 3464, and the status codes inside it in RFC 3463. A conforming DSN is a multipart/report message with a message/delivery-status part containing structured header blocks:

Reporting-MTA: dns; mail.sender-example.com
 
Original-Recipient: rfc822;[email protected]
Final-Recipient: rfc822;[email protected]
Action: failed
Status: 5.1.1
Remote-MTA: dns; mx.client-example.com
Diagnostic-Code: smtp; 550 5.1.1 <[email protected]>: Recipient address rejected: User unknown
Last-Attempt-Date: Tue, 27 Oct 2026 09:14:22 +0000

Key fields to read first:

FieldWhat it tells you
Actionfailed (final), delayed (still retrying), delivered, relayed, or expanded
StatusTransport-independent enhanced status code, e.g. 5.1.1
Final-RecipientThe address the server actually tried to deliver to
Remote-MTAThe server that rejected the message (absent = your own server detected the problem)
Diagnostic-CodeThe raw SMTP reply, usually the most precise detail

Two properties of DSNs explain many support tickets:

  1. A DSN is sent to the envelope sender (the SMTP MAIL FROM / Return-Path), not the visible From: header. That is why you can receive failure notices for mail that appears to come "from" you but that you never sent.
  2. A "failed" DSN is not absolute proof of non-delivery. RFC 3464 notes that SMTP race conditions can duplicate or lose delivery outcomes, so a message may occasionally reach the recipient despite a failure report.

Reading the Error: Enhanced Status Codes Explained#

Every DSN carries a three-part enhanced status code (x.y.z, RFC 3463). The first digit is the one to check before anything else:

  • 2.x.x — success. Rare in failure notifications.
  • 4.x.x — persistent temporary failure. The server keeps retrying for a while (commonly up to several days) before giving up. Do not "fix" anything permanent yet; the mail may still arrive.
  • 5.x.x — permanent failure. Retries will not help. The address, policy, or configuration must change.

The second digit narrows the area (1 = address, 2 = mailbox, 4 = routing/network, 5 = protocol, 7 = security or policy), and the third gives the precise condition. You will usually see the code paired with a plain SMTP reply such as:

550 5.1.1 <[email protected]>: Recipient address rejected: User unknown
421 4.7.0 Try again later, closing connection

The plain-language "550" or "421" is the SMTP reply code; the 5.1.1 / 4.7.0 is the enhanced status code. When both are present, quote the enhanced code — support teams and documentation are indexed by it.

Genuine Bounce or Forgery? Spotting Fake Failure Notifications#

Because DSNs are ordinary email, they can be forged as easily as any other message — RFC 3464 says so explicitly in its security considerations. Attackers exploit this in two well-known ways:

1. Fake "Delivery Status Notification (Failure)" phishing#

You receive a notice about a message "you sent" that failed, often with an HTML or archive attachment ("the original message", "details", "tracking info"). The attachment or a link leads to a credential-harvesting page or malware. Clues that the notice is fake:

  • You never sent the referenced message, and nothing matching it appears in your Sent folder.
  • The notice asks you to open an attachment or sign in to "release" the message. Legitimate DSNs return the original message as a passive message/rfc822 part and never demand credentials.
  • The From: display name says "Mail Delivery Subsystem" but the actual address is an unrelated lookalike domain.
  • The message fails authentication checks. Gmail provides message-authentication indicators and header details you can inspect; in Microsoft 365 headers, a Received-SPF value other than pass is a red flag worth investigating.

Treat unexpected failure notifications the way you treat any unsolicited attachment: do not open the attachment, do not click links, and verify through a channel you trust. Gmail's phishing guidance applies directly — never enter a password after clicking a link in a message, and use the built-in warnings and message-header checks.

2. Backscatter: real bounces for mail you never sent#

Sometimes the notification is genuine — but it reports a message you never sent. Spammers forge your address into the From: (and envelope sender) of their spam. When the spam bounces, the DSN is delivered to you. This is called backscatter.

How to tell backscatter from compromise:

  • The "original message" inside the DSN is spam you never wrote, often with a different Message-ID and recipients you do not know.
  • The Reporting-MTA / Remote-MTA fields name legitimate mail servers around the internet that rejected the spam.
  • Your account's Sent folder and login history show no corresponding activity.

What to do: you usually cannot stop spoofed mail at the receiving end, but domain owners can deploy SPF (RFC 7208), DKIM (RFC 6376), and DMARC (RFC 7489) with a reject policy so that receiving servers refuse mail that only claims to be from your domain. If bounces for unknown mail persist and your Sent folder does show messages you did not send, treat the account as compromised: change the password, revoke unknown sessions and app passwords, and check forwarding rules and automatic replies.

How to Triage Any Delivery Status Notification Failure#

Work top-down and the root cause usually surfaces within a minute:

  1. Confirm the notification is real. Check the sending server, authentication, and whether the referenced message exists in your Sent items (see previous section).
  2. Locate the failing recipient. Find Final-Recipient (or the "Delivery has failed to these recipients or groups" list). A multi-recipient send can succeed for some and fail for others — fix per recipient.
  3. Read Action. delayed means wait; failed means act.
  4. Decode Status. 4.x.x → temporary, retry later. 5.x.x → permanent, change something.
  5. Read Diagnostic-Code and the Remote-MTA. If a remote server appears, the recipient's side (or the path to it) rejected the message and you may need to contact them. If no remote server appears, your own sending server refused to transmit — the fix is on your side.
  6. Match the code to a cause using the table in the next section.

Common Causes and Fixes#

Bad or outdated recipient addresses#

Typical codes: 5.1.1 (bad destination mailbox address), 5.1.10 (recipient not found), 5.5.0 (mailbox unavailable at consumer providers)

The single most common delivery status notification failure error is simply an address that does not exist. Check for the usual suspects: typos (gmal.con), stray quotation marks, trailing dots after the address, and spaces pasted in from a document. If the address looks right, the mailbox may have been renamed or deleted — contact the recipient another way and update your address book. In Outlook, also clear the recipient from the auto-complete cache before resending: a stale cached address can keep failing even after you fix the message.

Spam and content policy rejections#

Typical codes: 5.7.1 (delivery not authorized / unable to relay), 5.7.23 (SPF violation), 5.7.26 (message must pass SPF or DKIM), 5.7.509 (DMARC reject policy), and Gmail's "flagged as spam or temporarily rejected" category

The receiving server accepted the connection but refused the message on policy grounds. Common triggers and fixes:

  • Links, attachments, or phrasing that trip spam filters — resend a plain-text version to test, and remove anything that requests personal information.
  • Sending one message to a very large recipient list — use a group or mailing-list service instead of a giant To/Cc header.
  • Missing or misaligned authentication — publish SPF and DKIM records for every sending domain and a DMARC record that tells receivers what to do. Google requires SPF or DKIM for all senders to Gmail, and SPF, DKIM, and DMARC for senders of more than 5,000 messages per day, with From: domain alignment for direct mail.
  • Your sending IP or domain is on a blocklist — the NDR usually names the list; delist after fixing the underlying cause (often a compromised account spamming through you).

Sending limits and rate quotas#

Typical codes: 5.2.2 (submission quota exceeded), 5.1.90 (daily recipient limit), 4.7.500-699 (suspicious activity, temporarily restricted)

Every provider throttles volume. Personal Gmail accounts, for example, stop at roughly 500 sent messages per day or 500 recipients on a single message, and a burst of undeliverable mail can also trigger a temporary block ("Messages you sent couldn't be delivered"). Microsoft 365 enforces per-sender, per-recipient, and per-tenant limits and reports them as 5.2.2, 5.2.121, 5.2.122, or 5.7.232–5.7.236. These errors are almost always temporary: pause sending, then resume at a lower rate within 24 hours. If you never send at those volumes, a limit error is a strong signal that your account has been hijacked and is being used for spam — reset credentials immediately.

Mailbox full, oversized messages, and temporary destination problems#

Typical codes: 4.4.7 (message expired), 4.4.316 (connection refused), 5.3.4 (message too big for system), generic 4.x.x, "out of storage space" notices

Three different destination-side problems look alike at first glance:

  • Mailbox over quota. The recipient must free storage before new mail can arrive.
  • Message exceeds size limits. Your provider may accept large attachments while the recipient's does not; a 5.3.4 status ("message too big for system", RFC 3463) means the message itself must shrink — host the files on a shared drive or file service and send a link instead.
  • Server or network outage. The destination server is down or overloaded, or a network path is broken. The sending server queues the message and retries; a delayed notification may arrive first, followed later by either delivery or a final failed report after the retry window (often several days).

Fixes: wait and resend later for transient outages; ask the recipient to free storage for quota errors; shrink or restructure oversized messages; and check the NDR for the remote server's IP so you can notify the party responsible when the problem is clearly on the receiving side.

Mail loops and runaway automatic replies#

Typical codes: 5.4.6 / 5.4.14 (routing loop detected)

When a message triggers an automatic response that itself triggers another response, mail servers start circling. Exchange Online breaks such loops after 20 iterations and issues an NDR; other servers do similar things. Loops typically come from misconfigured forwarding rules on one or both ends — for example, mailbox A forwards to mailbox B while both have out-of-office replies enabled. Check forwarding rules and automatic-reply settings on every account in the chain. Related hygiene, from RFC 3834: automatic responders should not answer messages that are themselves automated (they carry an Auto-Submitted header), should not respond to a null return-path (<>), and should not reply to addresses like MAILER-DAEMON — which is exactly the address legitimate DSNs come from.

Identity and protocol failures on the sending side#

Typical codes: 501 5.5.4 (invalid HELO/EHLO argument), 5.7.57 (client not authenticated at submission), 5.7.25 (IPv6 sender needs reverse DNS), 5.6.11 (invalid characters / bare line feeds)

Some failure notifications never leave your own infrastructure because the sending device is misconfigured:

  • Multifunction printers, scanners, and fax gateways often identify themselves with an empty or bogus HELO/EHLO value. SMTP (RFC 5321, section 4.1.1.1) requires the client to announce a fully qualified domain name or address; otherwise the server answers 501 5.5.4 and closes the connection. Fix the device's outbound mail profile (correct FQDN, working credentials, and a relay path your provider allows).
  • Unauthenticated submission fails when an application or device connects to a provider's SMTP endpoint without logging in. Configure authentication (modern deployments should prefer OAuth-based submission where supported).
  • Missing reverse DNS or broken TLS certificates surface as 5.7.25 (IPv6 sender without a reverse DNS record), 4.4.317 (remote certificate chain cannot be verified), or 4.7.322 (remote certificate expired). These need an administrator on the affected side, not a resend.

Provider Differences: Gmail vs. Microsoft 365#

Gmail. Bounces arrive from "Mail Delivery Subsystem" ([email protected]) with the subject line "Delivery Status Notification (Failure)". Gmail's help documentation maps the plain-language text in the body to fixes: address does not exist, message flagged as spam, sending limit reached, temporary recipient-side problem, recipient out of storage, and HELO/EHLO errors. The body explains what happened in plain English; the machine-readable status part is available in "Show original".

Microsoft 365 / Exchange / Outlook. The same message type is called a non-delivery report. Modern NDRs contain a plain-language cause, a "how to fix it" section for senders, and a "diagnostic information for administrators" block with the generating server, rejected recipient, remote server, enhanced status code, raw SMTP response, and original message headers. Useful header fields while troubleshooting are To, Received (the hop that produced the DSN), and Received-SPF. Microsoft also publishes an NDR diagnostic tool in the Microsoft 365 admin center: paste the code and it returns the cause and remediation.

Handling Failure Notifications Programmatically#

If you operate a web app, mailing list, or transactional mail service, DSNs arrive as multipart/report; report-type=delivery-status messages. Suppress addresses on Action: failed (and treat repeated Action: delayed as a warning), but do not remove a subscriber on a single report — even permanent-looking codes can be transient. A minimal parser using Python's standard library:

import email
 
with open("bounce.eml", "rb") as f:
    msg = email.message_from_binary_file(f)
 
for part in msg.walk():
    if part.get_content_type() == "message/delivery-status":
        for block in part.get_payload():
            if not block["Final-Recipient"]:
                continue  # per-message fields block (Reporting-MTA, ...)
            print("Recipient:", block["Final-Recipient"])
            print("Action:   ", block["Action"])
            print("Status:   ", block["Status"])
            print("Diagnostic:", block["Diagnostic-Code"])

The first header block in a message/delivery-status part holds per-message fields such as Reporting-MTA; each following block describes one recipient, which is why the loop skips blocks without a Final-Recipient field.

Practical rules for bounce handling:

  • Match bounces to submissions. Correlate Final-Recipient and Original-Recipient with your own send logs; never trust the visible From: header of the bounce.
  • Classify before acting. 5.1.x address errors → suppress and clean your list. 4.x.x → retry with backoff. 5.7.x policy errors → fix authentication or content before resending.
  • Keep lists clean. Sending repeatedly to invalid addresses is the fastest way to get your sending restricted at Gmail or throttled in Microsoft 365.
  • Automate, but verify. RFC 3834 warns responders not to answer DSNs themselves — an auto-reply to mailer-daemon can start a loop, and forged DSNs can trick naive automation into deleting valid subscribers.

Best Practices to Prevent Future Delivery Failures#

  1. Validate addresses at entry and re-verify stale lists before large sends. Correct typos before they become 5.1.1 bounces.
  2. Authenticate every sending domain with SPF, DKIM, and DMARC; monitor aggregate reports to catch lookalike senders impersonating you. This is also your main defense against the spoofed-mail backscatter that produces fake-looking failure notifications.
  3. Respect provider limits. Stay under personal Gmail's 500-messages-per-day ceiling on consumer accounts, use groups for bulk mail, and follow your provider's documented sending limits for business tiers.
  4. Send clean messages. One-click unsubscribe on bulk mail, valid From: formatting per RFC 5322, and no bare line-feed corruption (5.6.11) all reduce policy rejections.
  5. Configure devices properly. Give printers and scanners a real FQDN for HELO/EHLO, working credentials, and a supported relay path.
  6. Harden automatic replies. Disable out-of-office and auto-forwarding when they are not needed; when they are, make sure they cannot answer automated mail, and check forwarding chains for loops.
  7. Lock down accounts. A sudden flood of DSNs or an unexpected sending-limit error often means compromise — reset the password, revoke sessions, and review rules and connected apps.

Common Mistakes to Avoid#

  • Opening attachments in unexpected failure notifications. A real DSN returns the original message as inert text; anything that wants you to open an HTML page, macro document, or archive is a phishing attempt.
  • Resending unchanged to a permanent (5.x.x) error. The result is identical — fix the address, content, or configuration first.
  • Assuming "550" always means "the recipient blocked me." More often it is a nonexistent address, a quota, or an authentication gap; the enhanced status code distinguishes them.
  • Reading only the subject line. Gmail's "Delivery Status Notification (Failure)" is a generic label — the diagnosis is in the body and status fields.
  • Ignoring delay notifications. A 4.x.x DSN is an early warning; acting on it (contacting the recipient, freeing space) can prevent the eventual failure.
  • Deleting subscribers on the first bounce. Temporary outages produce real failure-looking codes; require persistent failure before suppression.

Conclusion#

The delivery status notification failure error is not a mystery — it is a structured report. Genuine notices carry RFC 3464 fields and an RFC 3463 status code that pinpoint the failing recipient and the responsible server, while forged ones rely on you panicking and opening the attachment. The reliable workflow is always the same: verify the notice is real, read Action and Status, match the code to a cause (bad address, policy rejection, quota, full mailbox, loop, or misconfigured sender), fix that cause, and only then resend. Pair that discipline with SPF, DKIM, and DMARC on your domains, clean recipient lists, and sane automatic-reply settings, and failure notifications become rare, quick to diagnose, and easy to resolve.

References#