Article
4xx vs 5xx SMTP Errors: What the Difference Means and What to Do
By Unlimited Verifier Team ·

Summary
A 4xx SMTP error is a temporary failure, so the sending server retries later, while a 5xx error is a permanent rejection that retrying will not fix. The first digit of the reply code determines the class, and the enhanced status code shows the specific cause so you can respond correctly.A 4xx SMTP error is a temporary failure, meaning the receiving server wants the sender to retry later, while a 5xx error is a permanent rejection that retrying won't fix. The first digit of the three-digit reply code tells you which one you have, and that digit should decide what your system does next.
Getting this wrong costs you in both directions. Retrying a 5xx wastes sending capacity and inflates your bounce count. Dropping an address after one 4xx can remove a real, engaged subscriber whose mail server was only busy. This guide explains how to read both classes, which codes you will actually see, and how to match each class to the right fix. For the full reference, see our bounce code reference.
Quick answer
- 4xx means retry. It is a temporary failure, and sending servers typically keep retrying for days before giving up.
- 5xx means stop. The message is rejected and returned. Retrying the same address won't change the result.
- The second and third digits name the cause.
550 5.1.1means the mailbox doesn't exist.554 5.7.1means a policy or reputation block. - Act on the enhanced code, not the class alone. 5.1.x errors are address problems that list cleaning fixes. 5.7.x errors are usually sender problems that authentication and list quality fix.
What is the difference between 4xx and 5xx SMTP errors?
A 4xx reply means "not now." A 5xx reply means "not for this message," and often "not for this address at all." The protocol defines that distinction, and the sending server's behaviour follows from it.
| 4xx: transient | 5xx: permanent | |
|---|---|---|
| Formal name in RFC 5321 | Transient negative completion reply | Permanent negative completion reply |
| Enhanced status class in RFC 3463 | 4.X.X, persistent transient failure | 5.X.X, permanent failure |
| What the sending server does | Keeps the message queued and retries | Stops and returns a bounce |
| How long it lasts | Until the sender's queue lifetime expires | Immediate; no retry |
| Typical causes | Throttling, greylisting, temporary mailbox problems | Unknown mailbox or domain, policy or reputation block, failed authentication |
| Usual ESP label | Soft bounce | Hard bounce, though policy blocks are labelled inconsistently |
| Right response | Back off and watch the trend | Fix the cause, then suppress dead addresses |
Bottom line: A 4xx asks for patience and a 5xx asks you to stop. Treat them as different events, not two severities of the same failure.
How do the digits in an SMTP code work?
Every SMTP reply starts with three digits. The first is the class. The second is a category: RFC 5321 assigns 0 to syntax, 1 to information, 2 to connections, 5 to the mail system, and leaves 3 and 4 unspecified. Many servers also append an enhanced status code in the form X.Y.Z, where X repeats the class, Y is the subject and Z is the detail.
The subject is where the practical information lives:
- 1: addressing (bad mailbox or domain)
- 2: mailbox status (full or disabled)
- 3: mail system
- 4: network and routing
- 5: mail delivery protocol
- 6: message content
- 7: security or policy
Two real decodes:
550 5.1.1: class 5 (permanent), subject 1 (addressing), detail 1 (bad destination mailbox). The address does not exist.421 4.7.0: class 4 (temporary), subject 7 (security or policy), detail 0 (unspecified). The server is slowing you down for a reason it doesn't explain.
Subject 7 causes the most confusion because it covers both temporary throttling (4.7.x) and permanent blocks (5.7.x). The class does the real work: a 4.7.x asks you to come back, while a 5.7.x says the message won't be accepted as sent. For a shorter reference, see the SMTP error code list.
Bottom line: Read the code as class plus subject. The subject tells you which fix applies, and the class tells you whether retrying is worth it.
What happens to a message after a 4xx error?
The sending server queues the message and retries on a schedule, usually with gaps that widen over time. The sender decides how long it keeps trying. Postfix holds deferred mail for 5 days by default before generating a bounce, according to its documentation. Other systems and hosted ESPs use their own windows, so there is no single industry figure.
The final bounce can mislead you. Under RFC 3464, the Action field says whether a recipient was delayed (still being retried) or failed (given up on). A failed delivery can still carry a 4.x status, such as 4.4.7 ("delivery time expired"). If you classify by the status code alone, you may treat a give-up as a temporary problem and keep mailing that address.
Throttling is the 4xx case most often mishandled. A 421 4.7.0 from a large provider usually signals that your sending IP or domain is sending too fast or has a weak reputation. The common mistake is retrying at full speed. The practical response is to lower concurrency per receiving domain and back off, because the throttle often stays in place under continued pressure.
Greylisting is different. It usually arrives as a 450 or 451 deferral with a 4.x.x code on first contact. It typically clears on a retry after a few minutes, because legitimate servers retry and many spam senders do not. A single 4xx from a new address is rarely a reason to suppress.
Bottom line: A 4xx is only temporary if it clears. Read the Action field, and treat a deferral that never clears as a failure.
What happens after a 5xx error?
The receiving server refuses the message, and the sending server generates a delivery status notification (DSN) reporting the failure. Under RFC 3464, that DSN carries Action: failed and a 5.x.x status. A correct sender stops there.
Two details matter in practice:
- Rejections are per recipient. SMTP answers each
RCPT TOseparately, so one batch of 1,000 recipients can produce many different outcomes. A single campaign can be partly delivered and partly rejected. - Rejections happen at different stages. A 5xx can arrive at connection time, before any content is sent, or after the message body. Early rejections usually reflect your sending identity or IP, not your copy, so rewriting the email won't fix them.
If DSNs keep arriving, the cause is usually on the sending side. Why MAILER-DAEMON messages keep arriving covers how to trace them.
Bottom line: A 5xx ends the attempt for that recipient. Your job is to find out whether the cause is the address or the sender.
Which 4xx and 5xx codes will you actually see?
The table covers the codes most often seen in sending logs. Provider wording changes, so build your reports on the code and subject rather than the text.
| Code | Class | What it usually means | Retry? | Fix |
|---|---|---|---|---|
421 4.7.0 |
4xx | Connection throttled or closed, often a rate or reputation signal | Yes, slowly | Lower concurrency per domain; see what 421 4.7.0 means |
450 4.2.1 |
4xx | Mailbox temporarily unavailable | Yes | Retry; suppress only if it persists past the queue lifetime |
450 or 451 with 4.x.x |
4xx | Greylisting on first contact | Yes | Usually clears on retry; don't suppress |
452 4.2.2 |
4xx | Mailbox full, reported as temporary by this server | Yes | Retry later and re-check |
550 5.1.1 |
5xx | User unknown; the address doesn't exist | No | Suppress immediately; see 550 5.1.1 explained |
550 5.1.2 |
5xx | Bad destination host or domain; no usable mail server | No | Suppress; see the 550 5.1.2 breakdown |
553 5.1.3 |
5xx | Address syntax is invalid | No | Fix at capture; remove malformed rows |
550 5.7.1 |
5xx | Policy or reputation block on the sender | No | Investigate the sender; don't suppress the whole list |
554 5.7.1 |
5xx | Rejected by policy, sometimes at connection time; see 554 5.7.1 policy blocks | No | Check authentication, complaint trend and blocklists |
550 5.7.26 |
5xx | Gmail blocks mail from an unauthenticated sender | No | Fix SPF and DKIM alignment |
550 5.7.515 |
5xx | Microsoft rejects mail that fails DMARC under a reject policy | No | Fix DMARC alignment |
For a shorter list, see 10 codes every email marketer should know.
Bottom line: Most rows point to one of two fixes: clean the address, or repair the sender.
Can a 4xx ever be permanent, and can a 5xx ever be temporary?
Yes. The class is the server's decision about whether it will accept the message later, not a fixed property of the problem.
Take a full mailbox. One server may return 452 4.2.2 and expect you to come back. Another may return 552 5.2.2, the RFC 5321 code for exceeded storage, and treat it as permanent. The condition is identical; the verdict differs.
Persistent 4xx errors also behave like blocks. A deferral that repeats for days from the same domain is functionally a rejection. Keeping that address in limbo means you keep paying for retries and distorting your delivery metrics.
The reverse happens too. Some 5.2.x mailbox errors clear within days. Treat a 5.2.x as hard only when it repeats across separate sends.
Bottom line: Judge the pattern over time. A single code records what one server decided once.
How should you classify 4xx and 5xx errors for suppression?
Suppression is where most list damage happens, in both directions. Use this framework on every send:
- Read the enhanced code and the DSN
Actionfield. The code gives the cause; the action tells you whether the sender has given up. - For 5.1.x (addressing), suppress on the first occurrence. The address is bad, and no retry will change that.
- For 5.2.x (mailbox), suppress after two occurrences across separate sends. Providers differ on whether a full mailbox is permanent, so the repeat test is a safeguard.
- For 5.7.x (security or policy), don't suppress by default. Check authentication, complaint trend and blocklist status first. Suppress only addresses that keep returning the same block after you fix the sender.
- For 4.x (transient), keep retrying on the sender's schedule. Suppress when the
Actionisfailed, or when deferrals persist across three separate sends or 72 hours. These two thresholds are our recommendation, not an industry standard. - Log every event with code, date, campaign and recipient domain, so patterns show up later.
Sources disagree here. Some platforms label every 5xx as a hard bounce, while others treat policy rejections as sender-side events. Where guidance conflicts, use the rule above. It separates address faults, fixed by cleaning the list, from sender faults, fixed by repairing the sending setup. For the categories themselves, see hard versus soft bounce codes and common soft bounce causes.
Bottom line: Suppress on evidence about the address, and fix the sender before blaming the list.
Why do 5.7.x errors need a different fix than 5.1.x?
Address errors are facts about a mailbox. Policy errors are judgments about the sender. A list can be spotless and still hit 5.7.1 walls, because the receiving server is scoring your sending identity, not your contacts.
Google's sender guidelines show the stakes. Google advises bulk senders to keep user-reported spam rates below 0.10% and to avoid reaching 0.30%. Those are complaint rates, not bounce rates, but complaints are a common trigger for policy blocks.
Use this diagnostic test. If your 5.1.x rate is low but you keep hitting 5.7.1 walls, stop cleaning and start checking the sender: SPF, DKIM and DMARC alignment, the complaint trend, and whether your sending domain or IP appears on a blocklist. Unlimited Verifier's free domain blacklist checker can help with the last step.
Verification supports email verification compliance and hygiene, but it can't lift a policy block. A cleaner list won't help if your authentication is failing.
Bottom line: 5.1.x is a list problem and 5.7.x is usually a sender problem. Don't use the wrong tool on either.
How does email verification reduce 5xx bounces, and what can't it do?
Verification runs checks that resemble what a receiving server does before accepting mail: syntax, domain, MX records and the SMTP mailbox. An address flagged as Invalid is the one most likely to produce a 5.1.x rejection. The advantage is timing. You catch the problem before the send, not in the bounce report afterward.
Catch-all domains need special handling. These domains accept mail for any local part at SMTP time, so a dead address on a catch-all domain may never produce a 5.1.1. Your bounce log can look clean for those domains even when the mailboxes are gone, so zero bounces there is weak evidence of validity. Unlimited Verifier detects catch-all domains and sorts results into Valid, Risky or Invalid, so don't read catch-all silence as proof. The catch-all email verification page explains how to handle them.
Verification has clear limits:
- It can't see 5.7.1 policy blocks.
- It can't prove consent.
- It can't predict future bounces for addresses that were valid when checked.
Teams running email verification for ecommerce and SaaS sign-up flows usually pair batch cleaning with real-time validation at signup, which keeps bad addresses out of the database altogether. For a broader view, see how to choose an email verifier and the best email verification tools roundup.
Unlimited Verifier's stated accuracy for deep verification is 99.5%. That is the company's own figure, not an independent benchmark. For automation, the email verification API accepts single checks or batches of up to 10,000 emails per request, and Pro includes 100,000 API emails a month. That covers the email verification API and automation workflow for teams that want checks before every send.
Bottom line: Verification removes the 5.1.x share of your bounces before you send. It does nothing for 5.7.x blocks.
How do you verify the 5.1.x addresses in a bounce log?
Follow these steps to turn a raw bounce report into a cleaned list:
- Export your bounce report from your ESP for the last 30 days. Include address, status code, date and campaign. If the platform doesn't offer a CSV, copy rows from the activity log.
- Sort rows into three files by enhanced code: 5.1.x, 4.x, and everything else. Record the count for each.
- Sign up for an account. The Free plan covers 25 lists a month of up to 150 emails each, so use Pro for larger files.
- Upload and verify a list with your 5.1.x file. CSV and Excel files up to 10MB are accepted, and the email column is detected automatically.
- Run verification. Pro handles lists of up to 25,000 addresses each, so split anything larger first.
- Read the three buckets. Invalid addresses get suppressed permanently. Risky addresses are held back. Valid addresses that appear in your 5.1.x file are a conflict: they bounced before but verify now. Don't re-include them until a small test send succeeds.
- Spot-check important addresses with the free email checker, especially high-value contacts that came back Risky.
- Export results as CSV, Excel or JSON before they expire. Pro keeps results for 30 days and Free keeps them for 24 hours. Save the export in your own records.
- Re-import and suppress. Load Valid contacts as a new segment in your ESP and add Invalid addresses to its suppression list. There are no native ESP, CRM or Zapier integrations, so this is a manual CSV import or an API script.
Bottom line: The 5.1.x file is the fastest win. Export, verify, suppress, and keep the counts.
What does a real bounce breakdown look like?
The numbers below are illustrative, not from a real send. They show how to split a campaign's outcomes by class and code.
Outcome of a 50,000-address campaign:
- Delivered (2xx, including 950 that succeeded on retry): 47,050 (94.1%)
- 4xx deferred, then bounced at queue expiry: 450 (0.9%)
- 5.1.1 unknown user: 1,650 (3.3%)
- 5.1.2 bad domain: 180 (0.4%)
- 5.7.1 policy block: 670 (1.3%)
Check: 47,050 + 450 + 1,650 + 180 + 670 = 50,000.
Analysis:
- Total bounce rate: 450 + 1,650 + 180 + 670 = 2,950, or 5.9% (2,950 ÷ 50,000). That is far above a 2% ceiling.
- Address-fixable bounces (5.1.x): 1,650 + 180 = 1,830, or 3.7% of the send. That is 73% of the 2,500 hard (5xx) bounces.
- Sender-side bounces (5.7.1): 670, or 1.3%. Cleaning the list won't touch these.
- Expired deferrals (4.x): 450, or 0.9%. These need a sending-speed or reputation fix.
What verification would have changed: Suppose the 1,830 bad addresses had been removed before sending. The list would shrink to 48,170, and the remaining bounces would be 670 + 450 = 1,120. That is 1,120 ÷ 48,170 = 2.3%, still above the 2% ceiling. The assumption is also generous, since no verifier catches every bad address. The leftover 1,120 bounces need sender and sending-speed fixes, not list work.
Cost and limits: The full 50,000-address list is above the Free plan's 150-email list cap, so Pro is required. It splits into two files of 25,000 each, within the Pro limit. Standard verification is unlimited on every plan with no per-email fee, so the cost is the Pro plan at $29 a month, or $25 a month billed annually. The email verification pricing comparison is simple here: $29 ÷ 50 = $0.58 per 1,000 addresses on monthly billing, and $25 ÷ 50 = $0.50 on annual. Pro also includes 25,000 deep verifications a month, separate from standard checks.
Read the example this way. The list problem was real, but it was only part of the bounce rate. Verification fixes the bucket it can see, and the sender-side buckets need separate work.
Bottom line: Split bounces by code before deciding what to fix. Here, verification addresses 1,830 of 2,950 bounces, and the other 1,120 need different work.
What mistakes keep bounce rates high?
- Retrying 5xx errors "just in case." Each retry to a 5.1.1 address adds a bounce and signals poor list hygiene.
- Suppressing every 5.7.1. You remove engaged subscribers blocked by your sending identity, and the block continues.
- Trusting the ESP's hard or soft label without reading the code. Labels summarise. The enhanced code is the evidence.
- Ignoring the DSN
Actionfield. Afailedwith a 4.x status is a give-up, not a temporary delay. - Reading zero bounces at a catch-all domain as proof of validity. Those servers accept mail at SMTP time, so dead mailboxes stay silent.
- Letting verification results expire. Free keeps them for 24 hours and Pro for 30 days. Export and archive.
- Counting a 250 accept as inbox delivery. A 2xx means the receiving server accepted the message, not that it reached the inbox.
- Verifying once and never again. Addresses decay, so schedule re-verification. The list maintenance best practices guide covers a schedule.
- Using verification to fix a 5.7.1 block. It can't, and time spent cleaning is time not spent on authentication and complaint rates.
Bottom line: Most mistakes come from treating every failure as the same kind of failure.
How we compared (and limitations)
This guide is based on public documentation: the reply-code definitions in RFC 5321, the enhanced status codes in RFC 3463, the bounce report format in RFC 3464, Postfix's documented default queue lifetime, and Google's published sender guidelines.
We did not run controlled sends to mailbox providers, so the provider-specific examples describe commonly documented behaviour. Those texts and policies change without notice. The 99.5% accuracy figure is Unlimited Verifier's own stated number for deep verification, not an independent benchmark. The worked example is illustrative, and the retry and suppression thresholds are our recommendations, not industry standards. Unlimited Verifier's verification engine is published as open source on GitHub, so you can inspect how the SMTP check is implemented.
Bottom line: Use the RFC definitions as the stable reference, and check provider wording against your own bounce logs.
Which fix matches which code
| Code | Class | Meaning | Action |
|---|---|---|---|
| 421 4.7.0 | 4xx | Temporary policy or rate limit | Back off and retry |
| 450 4.2.2 | 4xx | Mailbox full, temporary | Retry, watch trend |
| 550 5.1.1 | 5xx | Mailbox does not exist | Suppress address |
| 554 5.7.1 | 5xx | Policy or reputation block | Check authentication and list quality |
Frequently asked questions
What does a 4xx SMTP error mean?
A 4xx SMTP error is a temporary failure. The receiving server is asking the sender to try again later, usually because of throttling, greylisting, or a temporarily full or busy mailbox. Sending servers normally keep the message queued and retry for a period before giving up.
What does a 5xx SMTP error mean?
A 5xx SMTP error is a permanent rejection. The receiving server will not accept the message, and retrying the same address will not change the result. Common causes include a nonexistent mailbox or domain, or a policy or reputation block.
Should I retry a 5xx error?
No. A 5xx code signals a permanent failure, so retrying wastes sending capacity and inflates bounce counts. Fix the underlying cause if it is on your side, such as authentication or content, and suppress addresses that are genuinely invalid.
How long does a sender retry a 4xx error?
It depends on the sending server's queue lifetime. Many mail servers keep retrying for several days before returning a bounce, though the exact window is set by the sender's configuration and is not a universal standard.
What is an enhanced SMTP status code?
An enhanced status code is a three-part code in the form X.Y.Z that follows the basic reply code, defined in RFC 3463. The first part repeats the class, the second names the subject, and the third gives detail on the specific cause.
What does 550 5.1.1 mean?
550 5.1.1 is a permanent failure indicating that the destination mailbox does not exist. It is an addressing problem, so the fix is to remove or correct the address on your list rather than retry.
What does 421 4.7.0 mean?
421 4.7.0 is a temporary security or policy response. The server is refusing the connection or message for now, often due to rate limits or reputation checks, so retrying later with a healthy sending pattern is the usual approach.
Why does my ESP label some 5xx errors as soft bounces?
Bounce labelling is not standardised across providers. Some ESPs classify certain policy-based 5xx responses as soft bounces, so check the raw SMTP reply and enhanced code instead of relying only on the platform label.
Is a 5xx error always caused by a bad email address?
No. While 5.1.x codes point to address problems, 5.7.x codes usually reflect sender-side issues such as failed authentication, poor list quality, or reputation blocks.
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.