Article
10 SMTP Error Codes Every Email Marketer Should Know (With Actions)
By Unlimited Verifier Team ·

Summary
The SMTP error codes that matter most for email marketers are 550 5.1.1, 550 5.1.2, 553 5.1.3, 554 5.7.1 and six others covering full mailboxes, throttling, deferrals, access blocks and login failures. Each code calls for a different action: suppress permanent recipient failures, correct malformed addresses, retry temporary deferrals, and investigate policy refusals.The ten SMTP error codes every email marketer should know are 550 5.1.1 and 550 5.1.2 (permanent recipient or domain failures), 553 5.1.3 (malformed address), 554 5.7.1 (policy refusal), and six more covering full mailboxes, throttling, temporary deferrals, access blocks and login failures. Each one calls for a different action, and 554 5.7.1 is the one most often misread. This guide covers all ten for list owners, agencies and SaaS teams cleaning large databases for deliverability and compliance. For the wider framework, start with our email bounce codes guide.
Quick answer
- Suppress: 550 5.1.1 (mailbox doesn't exist) and 550 5.1.2 (domain can't receive mail) are permanent recipient failures. Suppress after the first one.
- Fix at the source: 553 5.1.3 means the address is malformed. Correct obvious errors and re-check, or drop it.
- Retry: 450 4.2.1 and 450 4.7.1 are temporary. Judge them by the final outcome after the retry window.
- Investigate, don't suppress: 554 5.7.1 and 535 5.7.8 usually point to your sending IP, content, authentication or credentials.
- Benchmark: keep hard bounces under 1% of each send. This is our recommended threshold, not an industry standard.
Which SMTP error codes matter most for email marketers?
Read three parts of every bounce: the reply code, the enhanced code and the text. The first digit of the reply code is the most important signal. Under RFC 5321, a 4xx reply is temporary and a 5xx reply is permanent. The enhanced code, such as 5.1.1, adds the subject: 5 is permanent and 4 is transient, and the middle number identifies the category (addressing, mailbox, routing, policy or security). For a deeper walkthrough, see 4xx vs. 5xx SMTP errors and the enhanced status code list.
Quick picks by situation
- Cleaning an imported or aging list: 550 5.1.1, 550 5.1.2, 553 5.1.3
- Diagnosing a bounce spike: 554 5.7.1, 421 4.7.0, 535 5.7.8
- Explaining delayed mail: 450 4.7.1, 450 4.2.1
- Deciding what to do with full inboxes: 552 5.2.2
- Working out a domain-wide rejection: 550 5.4.1
| Code | What it means | Hard or soft | First action |
|---|---|---|---|
| 550 5.1.1 | Mailbox doesn't exist | Hard | Suppress |
| 550 5.1.2 | Domain or host can't be reached | Hard (domain-level) | Confirm mail server, suppress domain |
| 553 5.1.3 | Address syntax invalid | Hard | Correct or suppress |
| 552 5.2.2 | Mailbox full | Permanent code; often treated as soft | Retry, then apply your written rule |
| 450 4.2.1 | Mailbox temporarily not accepting mail | Soft | Retry |
| 450 4.7.1 | Greylisting or temporary policy deferral | Soft | Retry |
| 421 4.7.0 | Throttled or reputation-limited | Soft (warning if persistent) | Slow down, check reputation |
| 554 5.7.1 | Refused for policy or blocking | Policy (address may be valid) | Investigate sender |
| 535 5.7.8 | Login credentials rejected | Sender-side | Fix credentials |
| 550 5.4.1 | "Access denied" for recipient | Usually hard; wording varies | Check single addresses vs. whole domain |
Bottom line: the reply code sets the timing, the enhanced code sets the subject, and the table above is a default to test against your own logs.
How we compared these codes (and limitations)
This comparison is based on public documentation: the SMTP standards and common provider behavior. We judged each code on four criteria:
- Frequency in marketing sends: codes that commonly appear in ESP bounce reports and list-cleaning work.
- Decision difference: codes where the right action is clear and where marketers often act incorrectly.
- Standards grounding: each code is defined in RFC 5321 (reply codes), RFC 3463 (enhanced status codes) or RFC 4954 (SMTP authentication), or is documented by a major mailbox provider.
- Coverage: recipient, domain, policy and sender-side failures, so no single category dominates the list.
Limitations. We did not analyze a bounce dataset, so we make no claim about how often each code appears on your list. Receiving servers word messages differently and sometimes omit enhanced codes. The same code can mean different things at different providers, and vendors don't always use enhanced codes as the RFCs define them. Thresholds such as the 1% hard-bounce target are recommendations, because no single number is standardized across providers. If you use a verification tool, its accuracy figures are the vendor's claims. For example, Unlimited Verifier states 99.5% accuracy for deep verification; test any claim on a sample of your own list.
Bottom line: the codes and enhanced codes tell you more than the text, but neither replaces reading your own bounce logs.
What are the 10 SMTP error codes every email marketer should know?
Each entry covers what the code means, whether it's hard or soft, what usually triggers it, what to do, and where marketers go wrong.
1. 550 5.1.1: Mailbox doesn't exist
- What it means: the receiving server says no mailbox exists for that address at that domain.
- Hard or soft: hard (permanent).
- Common triggers: typos, former employees whose mailboxes were deleted, closed consumer accounts, and addresses collected from paper forms or old lists.
- What to do: suppress the address after the first 5.1.1. Retrying a nonexistent mailbox adds bounce history with no chance of delivery.
- Practitioner point: if several people at one company bounce 5.1.1 on the same send, the organization may have changed its address format (for example, moving from first.last to flast) or migrated mail systems. Test a known-good address at that domain before suppressing them all.
- Where it's misread: marketers often retry 5.1.1 addresses "to be safe." That doesn't help and makes your bounce history worse.
For the full breakdown, see 550 5.1.1 user unknown meaning.
2. 550 5.1.2: Bad destination host or domain
- What it means: the sending server can't find a usable mail server for the domain. The problem is the part after the @.
- Hard or soft: hard, at the domain level.
- Common triggers: a lapsed domain, a company that retired its domain after a merger, a domain that no longer runs mail, or a misspelling such as gmial.com or yaho.com.
- What to do: run an MX lookup. If the domain has no working mail server, suppress every address at that domain, not just the one that bounced. Fix a misspelled provider domain only if the subscriber confirms the corrected address.
- Practitioner point: one 5.1.2 usually predicts more of the same. Repeat sends to a dead domain generate hard bounces across an entire segment, so domain-level suppression is faster than address-by-address cleanup.
- Where it's misread: treating 5.1.2 as an individual address problem and retrying other addresses on the same domain.
See 550 5.1.2 bad destination host meaning for more on domain-level failures.
3. 553 5.1.3: Invalid address syntax
- What it means: the address breaks the format the server accepts.
- Hard or soft: hard, but usually fixable.
- Common triggers: spaces inside the address, a doubled @, a trailing comma or semicolon from a spreadsheet export, stray quotation marks, or a pasted "mailto:" prefix.
- What to do: trim whitespace and punctuation, correct obvious typos, re-verify, then send. Anything still malformed gets suppressed.
- Practitioner point: most syntax failures enter through signup forms and CSV imports, so the cheapest fix happens at capture. Real-time validation at signup prevents most of them. Some servers reject internationalized (non-ASCII) addresses that lack SMTPUTF8 support, defined in RFC 6531. Confirm a 553 before assuming the address is fake.
- Where it's misread: assuming every 553 is a fake address. Some are valid addresses your server or the recipient's server can't parse.
4. 552 5.2.2: Mailbox full
- What it means: the recipient's storage quota is exhausted. The 552 reply is a permanent 5xx code, and RFC 3463 defines 5.2.2 as a full mailbox.
- Hard or soft: permanent by the standard, but commonly treated as soft because full inboxes often clear within days.
- Common triggers: inboxes with no quota management, dormant accounts that no one cleans out, and shared or legacy mailboxes.
- What to do: retry on your normal schedule. If the address returns 5.2.2 across several sends, move it to suppression or a re-engagement track under a written rule. For example, you might define "three sends within 30 days" as the trigger. That number is a suggested policy, not a standard.
- Practitioner point: the standard and common ESP practice disagree here. Pick one rule, write it down and apply it to every list.
- Where it's misread: treating every 5xx as automatically permanent, which removes engaged subscribers whose inboxes simply filled up.
Related reading: common soft bounce reasons.
5. 450 4.2.1: Mailbox temporarily unavailable
- What it means: RFC 3463 defines 4.2.1 as a mailbox that is disabled and not accepting messages at the moment. The 4xx reply tells the sender to try again.
- Hard or soft: soft.
- Common triggers: accounts temporarily disabled by the provider, maintenance on the recipient's mail system, and mailbox locks during migrations.
- What to do: let your sending platform retry. Treat the address as undeliverable only after it fails across the full retry window.
- Practitioner point: a sudden burst of 4.2.1 across many unrelated domains often points to a provider-side incident, not your list. Check whether it clears within a day before taking any list action.
- Where it's misread: counting the first deferral as a final hard bounce, which inflates your hard-bounce rate.
See 450 requested mail action not taken meaning for the full set of 450 variants.
6. 450 4.7.1: Greylisting or temporary policy deferral
- What it means: the receiving server temporarily refuses mail from an unfamiliar sender and recipient pair (greylisting), or applies a policy deferral, expecting a legitimate sender to retry later.
- Hard or soft: soft. A retry usually succeeds.
- Common triggers: first contact from a new sending IP, a sudden volume spike, or a policy check that requires a retry.
- What to do: confirm your platform retries deferred mail and that it doesn't count the first deferral as a bounce.
- Practitioner point: a campaign full of 4.7.1 looks alarming in real time but usually resolves on retry. Judge the send by its final status after the retry window, not the first hour. RFC 5321 describes sending servers that keep retrying for days, not minutes, before giving up.
- Where it's misread: cutting or cleaning a list because of a deferral wave that would have cleared on its own.
7. 421 4.7.0: Throttled or reputation-limited
- What it means: the server asks you to slow down because connection or sending rate is too high, or because your sending reputation is under review. The server closes the connection.
- Hard or soft: soft, but a warning sign when it persists.
- Common triggers: too many simultaneous connections to one provider, bursts of sends, and high complaint or low engagement rates among that provider's users.
- What to do: reduce concurrency to that provider and pause if deferrals continue over several days. Google's sender guidelines describe keeping the spam rate in Postmaster Tools below 0.3% and ideally under 0.1%. Google's own messages are the best guide to what it's signaling.
- Practitioner point: scattered deferrals are noise. Deferrals that climb over several days at one provider signal reputation. Retrying harder makes it worse, so clean the list and reduce volume first.
- Where it's misread: treating 421 as a list-quality problem when it's really a volume or reputation problem.
See 421 4.7.0 try again later meaning for the full explanation.
8. 554 5.7.1: Delivery not authorized, message refused
- What it means: the receiving system refused the message for policy reasons. RFC 3463 defines 5.7.1 as delivery not authorized. Causes include a listed sending IP, failed authentication, or content that trips filters.
- Hard or soft: a policy rejection. The recipient address may be perfectly valid.
- Common triggers: a sending IP on a blocklist, SPF, DKIM or DMARC failures, reputation-based refusals, and content filters.
- What to do: don't suppress recipients on this code alone. Check your sending IP and domain against blocklists, confirm that authentication passes, and compare refusals across providers.
- Practitioner point: 554 5.7.1 is the code most often misread as a bad address. Suppressing these recipients shrinks your list while the real cause keeps hurting every send. If refusals cluster on one provider, look at your reputation there. If they appear across many providers at once, look at your sending IP or content.
- Where it's misread: suppressing thousands of valid subscribers and wondering why the problem persists.
See what does 554 5.7.1 mean for detail on the common variants.
9. 535 5.7.8: Authentication credentials invalid
- What it means: your SMTP login was rejected. RFC 4954 defines SMTP AUTH and the 535 reply for failed authentication, and RFC 3463 defines 5.7.8 as invalid credentials.
- Hard or soft: a sender-side configuration error.
- Common triggers: a wrong password, a rotated or revoked API key, an app-specific password that was never created, and a deleted SMTP user.
- What to do: fix the credentials in your sending tool, then resend the affected batch. Never suppress recipients for this code.
- Practitioner point: a failed login fails an entire batch at once, so it can look like a list disaster on a dashboard. Check your authentication settings before touching the list. A 535 that appears on every batch right after a password change is a configuration issue, not a list issue.
- Where it's misread: blaming the list for a broken login and suppressing recipients who did nothing wrong.
For the wider set of login and relay errors, see the SMTP code reference list.
10. 550 5.4.1: Recipient address rejected, access denied
- What it means: the wording is commonly reported from Microsoft 365 and Exchange Online, where the destination's directory-based blocking rejects recipients it doesn't recognize. Other systems use similar text for policy blocks.
- Hard or soft: usually hard, but the wording is ambiguous.
- Common triggers: addresses that don't exist in the destination directory, and domain-wide policies that block inbound mail from your sending source.
- What to do: count how many addresses at the domain return it on the same send. A few single addresses usually mean nonexistent recipients, so suppress those. If a whole domain returns it, treat it as a policy event: verify a sample of addresses and check with the recipient before suppressing the domain.
- Practitioner point: match on the code and the pattern, not the words alone. A domain-wide 5.4.1 is a policy event, not a list-quality event.
- Where it's misread: assuming every 550 means the mailbox is gone, then deleting an entire customer domain that was only temporarily blocked.
Bottom line: the first digit sets whether to retry, the enhanced code sets the subject, and the pattern across addresses sets the action.
How do you triage bounce codes into a clean list with Unlimited Verifier?
Run this triage after any bounce spike and as a monthly habit. It needs only your ESP's bounce export, a spreadsheet and an account. If you don't have one yet, you can sign up for a free account. Free covers small lists.
- Export the bounce report from your ESP as CSV, keeping the address, bounce text and any code column. If there's no code column, filter the text for codes beginning with "5." or "4.".
- Group by code. Sort into three piles: 5.1.x and 552 5.2.2 (hard candidates), 4.x codes (soft), and 554 5.7.1, 535 5.7.8 and 550 5.4.1 (investigate before suppressing).
- Build a hard-bounce file. Keep only the confirmed 5.1.x addresses in one column named "email."
- Upload and verify the file. The email column is detected automatically. Files can be up to 10MB. On Free, keep each list to 150 emails. On Pro, keep each list to 25,000.
- Read the buckets. Invalid confirms suppression. Valid on an address that just bounced is a conflict to investigate, not a reason to mail it. Risky addresses need segmenting, and any catch-all, disposable or role-based flags tell you why.
- Download immediately. Export as CSV, Excel or JSON. Free results are kept 24 hours and Pro results are kept 30 days, so archive every export.
- Re-import. Add Invalid addresses to your ESP's suppression list, tag Risky, and keep the soft-bounce pile on its retry schedule. If you're unsure about a single suspicious address, run it through the free email checker first.
- Verify the rest of the list. Bounced addresses are already known bad. The value is catching the next bad ones before they bounce.
- Automate later. On Pro, you can generate API keys for single checks and for batches of up to 10,000 emails per request, within 100,000 emails a month. This is the core of email verification API and automation.
Worked example
Take a 40,000-address list. The last send produced 900 × 550 5.1.1, 150 × 550 5.1.2, 60 × 553 5.1.3 and 90 × 552 5.2.2.
- Hard-bounce rate: 900 + 150 + 60 = 1,110. Then 1,110 ÷ 40,000 = 2.78%. The 1% target allows 400 hard bounces, so this send produced about 2.8 times that limit.
- Suppress now: the 1,110 hard bounces go on your suppression list immediately.
- Hold the 552s: the 90 × 5.2.2 addresses stay out of the hard count until your retry rule decides them.
- Verify the rest: 40,000 − 1,110 = 38,890 addresses. On Pro, that splits into 25,000 + 13,890 across two lists, both within the 25,000-per-list cap.
- On Free: 38,890 ÷ 150 = 260 lists. The Free plan allows 25 lists a month, or 3,750 emails. At that cap, 38,890 ÷ 3,750 ≈ 10.4 months. Free works for triaging a small hard-bounce file, but not for a full-list pass on a 40,000-address list.
- Cost: Pro is $25/month billed annually, or $300 a year. For a clear view of how verification costs compare across providers, see email verification pricing.
The next send's bounce rate depends on how many of the remaining 38,890 addresses are invalid. The verification run is how you measure that.
Bottom line: suppress confirmed hard bounces first, then verify the rest of the list so the next send lands under 1%.
How do you choose the right action for a bounce code?
Match the pattern first, then the code. The same code can call for different actions depending on whether it hits one address, one domain or many domains.
| Pattern | Likely cause | Action |
|---|---|---|
| One address, 550 5.1.1 or 553 5.1.3 | Nonexistent or malformed address | Suppress or correct |
| One domain, 550 5.1.2 | Dead or misspelled domain | Confirm mail server, suppress domain |
| One address, 552 5.2.2 across sends | Persistently full mailbox | Retry, then apply your written rule |
| Many domains, 4.x codes | Provider incident or throttling | Retry; check whether it clears |
| One provider, rising 421 4.7.0 | Reputation or volume | Cut volume, review list and authentication |
| 554 5.7.1 across domains | Your IP, domain or content | Investigate sender; don't suppress |
| 535 5.7.8 on the whole batch | Broken credentials | Fix configuration |
| 550 5.4.1 on a few addresses | Nonexistent recipients | Suppress those addresses |
| 550 5.4.1 across one domain | Domain-wide policy | Verify a sample, check before suppressing the domain |
If you're choosing a tool to run this triage, see how to choose an email verifier and our overview of best email verification tools. For SaaS signup flows where bad addresses enter at capture, see email verification for ecommerce and SaaS.
Bottom line: a single address points to the recipient, a single domain points to the domain, and many domains point to you.
What mistakes do marketers make with SMTP codes?
- Reading only the first digit. 552 is a 5xx code but is often treated as soft. 421 is a 4xx code that can signal reputation trouble. Read the enhanced code too.
- Suppressing 554 5.7.1 recipients. You shrink the list and leave the real reputation problem untouched.
- Retrying 550 5.1.1 addresses. Repeat attempts add bounce history without any chance of delivery.
- Counting 4xx deferrals as bounces. Temporary failures inflate your hard-bounce rate and can trigger unnecessary list cleaning.
- Acting on one send. A single spike at one provider can be an incident. Wait for a second send or a day of data before making list decisions.
- Treating a 250 response as proof. Catch-all domains accept mail for any address during the SMTP exchange, so acceptance doesn't confirm the mailbox. See Catch-all email verification for how catch-all domains are detected.
- Forgetting to re-import suppressions. Otherwise the next import brings dead addresses back. Suppression is part of email verification compliance and hygiene.
- Assuming bounce notices mean your list is bad. Bounce messages for mail you never sent often come from forged sender addresses. Those aren't list problems. See why you keep getting mailer-daemon bounce messages.
- Matching on wording. Providers reword messages, so match on the code, the enhanced code and the pattern together.
Bottom line: most bounce mistakes come from acting on text or the first digit alone, instead of reading the code, subject and pattern together.
Verdict
The ten codes reduce to three actions. Suppress permanent recipient failures (550 5.1.1, 550 5.1.2, 553 5.1.3 once it's confirmed malformed). Retry temporary ones (450 4.2.1, 450 4.7.1, and 552 5.2.2 under a written rule). Investigate sender-side and policy codes (554 5.7.1, 535 5.7.8, a persistent 421 4.7.0, and domain-wide 550 5.4.1).
Verification handles the first group before the damage happens. Unlimited Verifier runs syntax, domain, MX and SMTP mailbox checks, sorts results into Valid, Risky and Invalid, and offers unlimited standard verification on every plan, including Free. The trade-off is that there are no native ESP, CRM or Zapier integrations, so the workflow is export, verify and re-import, or an API call on Pro. For most list owners, that's a 15-minute routine once a week.
Bottom line: suppress the permanent failures, retry the temporary ones, and investigate anything that points at your own sending setup.
Bounce response playbook
| Code | Decision | Action |
|---|---|---|
| 550 5.1.1 | Suppress | Remove after first occurrence |
| 553 5.1.3 | Fix or suppress | Correct obvious typos, else remove |
| 450 4.2.1 | Retry | Retry within the window, then apply your rule |
| 554 5.7.1 | Investigate | Check IP, content, authentication and complaints |
| 535 5.7.8 | Fix credentials | Check login details, do not touch the list |
Frequently asked questions
What is the difference between a hard bounce and a soft bounce?
A hard bounce is a permanent failure, usually a 5xx reply such as 550 5.1.1, meaning the address should be removed. A soft bounce is a temporary failure, usually a 4xx reply such as 450 4.2.1, meaning the message can be retried later. Many ESPs classify bounces this way, but some treat certain 5xx codes such as 552 5.2.2 as soft in practice, so check your provider's rules.
Should I suppress an address after one 550 5.1.1 bounce?
Yes. A 550 5.1.1 response means the mailbox does not exist, which is a permanent recipient failure. Retrying will not fix it, and repeated attempts can hurt your sender reputation. Suppress the address after the first occurrence and keep it off future sends unless the subscriber re-confirms.
What does 554 5.7.1 mean?
554 5.7.1 is a policy refusal. The receiving server rejected the message because of its policy, such as sender reputation, content, authentication or IP blocking. It does not necessarily mean the recipient address is invalid. Investigate your sending setup, authentication records and complaint or bounce trends before changing your list.
How long should I retry a soft bounce?
Retry soft bounces for the retry window your sending platform or MTA defines, then judge the message by its final outcome. Most platforms queue and retry temporary failures over a period of hours to days before giving up. If the address still fails after the window ends, apply your written rule for soft bounces, such as suppression after several consecutive failures.
Does 421 4.7.0 mean I have been blocked?
Not necessarily. 421 4.7.0 is a temporary response indicating the server is throttling your connection or limiting you based on reputation. It is a signal to slow down. If it persists across sends, treat it as a warning, reduce volume and rate, and check your sending reputation and authentication.
What does 535 5.7.8 mean for email marketers?
535 5.7.8 means the server rejected your login credentials during SMTP authentication. This is a sender-side problem, not a recipient problem. Check the username, password or API key, confirm the authentication method your provider requires, and make sure the credentials have not been rotated or revoked. Do not remove recipients based on this code.
What does 550 5.4.1 mean?
550 5.4.1 is generally an access-denied response, often seen when a receiving system refuses delivery. The wording varies by provider, so it can mean a single address is blocked or that the whole domain rejects you. Test a few known addresses at the same domain. If many addresses at one domain return it, treat it as a domain-level issue and investigate before suppressing individually.
What is an enhanced SMTP status code?
An enhanced status code, such as 5.1.1, is a three-part code defined in RFC 3463 that adds detail to the basic SMTP reply. The first number is the class: 2 is success, 4 is persistent transient failure and 5 is permanent failure. The second number is the subject, such as 1 for addressing or 7 for security or policy. The third number gives the specific detail.
Should I remove addresses that return 553 5.1.3?
Correct the address if the error is obvious, such as a missing @ sign, a typo in the domain or a stray space, and then re-verify it. If the address cannot be corrected with confidence, suppress it. A 553 5.1.3 code means the address syntax is invalid, so it will never deliver as written and should not stay in an active sending list.
Cut Bounces and Protect Your Sender Reputation With Unlimited Email Verification
Every invalid address you send to hurts your sender reputation and your inbox placement. Verifying your list before each send means fewer bounces, healthier deliverability, and a predictable cost. With Unlimited Verifier, standard verification is unlimited on every plan, with no per-email fees.
Unlimited Verifier checks syntax, domain, MX records and the SMTP mailbox. It also detects catch-all domains and flags disposable and role-based addresses. Export a CSV from your email platform or CRM, upload and verify the list, then re-import the clean results. You can also automate checks with the email verification API.
| Free | Pro | |
|---|---|---|
| Price | $0 | $29/month ($25/month billed annually) |
| Lists | 25 a month, up to 150 emails each | Unlimited, up to 25,000 emails each |
| Deep verifications | 750 a month | 25,000 a month |
| API access | No | Yes (100,000 emails a month) |
| Result retention | 24 hours | 30 days |
Why teams choose Unlimited Verifier
- Unlimited standard verification on all plans, including Free, with no per-email fees.
- Catch-all detection so you know which domains accept every address.
- Clear Valid / Risky / Invalid results, exportable as CSV, Excel or JSON.
- An open-source verification engine you can inspect on GitHub.
Unlimited Verifier states 99.5% accuracy for deep verification.
Create a free account and verify your next list before you hit send.
Want to test a single address first? Try the free email checker.