Article

Why Do Mail Servers Block Email Verification Probes?

By Unlimited Verifier Team ·

Diagram showing a mail verification probe sending RCPT TO to a mail server, which replies with accept, invalid, or throttle codes.

Summary

Mail servers block email verification probes mainly to stop directory harvest attacks, where senders guess many addresses at one domain using RCPT TO. Other triggers include poor IP reputation, blocked port 25, and misconfigured DNS. Blocked probes should be recorded as unknown rather than invalid.

Mail servers block email verification probes mainly to stop directory harvest attacks, in which senders guess thousands of addresses at one domain using RCPT TO commands. They also refuse connections from poor-reputation IPs, blocked port 25 and misconfigured DNS. A verifier that hits these defences gets no mailbox answer, so it should record "unknown," not "invalid."

Blocking is rarely a verdict on an individual address. The receiving server is reacting to the pattern and source of your connection, and that distinction is the key to reading any verification result correctly.

Quick answer

What does a verification probe send to a mail server?

A probe connects to the domain's mail exchanger on TCP port 25 and runs a short SMTP dialogue: EHLO, MAIL FROM, RCPT TO, then QUIT. It never sends DATA, so no message body is transmitted. The reply to RCPT TO is the only mailbox signal the verifier receives. SMTP is defined in RFC 5321, and the enhanced status code after each reply is defined in RFC 3463.

Before any probe runs, the verifier checks syntax and whether the domain accepts mail, so only plausible addresses reach a server. Most of the trouble starts at the probe itself. This mapping shows how to read each reply:

Reply What it means How to interpret it
250 2.1.5 Recipient accepted Mailbox probably exists, unless the domain is catch-all
550 5.1.1 Mailbox unknown (permanent) Invalid, provided the text is not a policy message
550 5.7.x Refused on security or policy grounds A signal about your connection, not about the address
4xx with x.7.x (e.g., 421 4.7.0) Temporary throttling or greylisting Retry later; do not mark Invalid
No reply or timeout Connection dropped or deliberately slowed Unknown; retry

The enhanced code is the detail most people skip. RFC 3463 places x.7.x codes in the security and policy class, so a 5.7 or 4.7 reply describes your connection or sending behaviour, not the mailbox.

Bottom line: Only a clean 250 or a mailbox-specific 550 5.1.1 says anything about the address. A 5.7.x reply, a 4xx reply or silence describes how the server responded to you.

Why do mail servers block verification probes?

Five recurring causes explain most blocked probes. They look alike in a log file, so identify which one you face before changing anything.

1. Directory harvest attacks. Harvesters try thousands of guessed addresses at one domain in a short window. Servers defend by rate-limiting per IP, adding delays, disabling VRFY (which RFC 5321 notes can be abused for harvesting), or accepting every recipient so guesses reveal nothing. The server cannot see your intent. It sees a burst of RCPT TO commands against one domain, which matches a harvester's pattern. For example, 40,000 RCPT TO commands against one domain in twenty minutes will trigger defences even if every address is valid.

2. IP reputation and policy lists. Spamhaus publishes a Policy Block List (PBL) covering IP ranges that should not deliver mail directly to a mail exchanger, typically dynamic or residential space. A probe run from a home or broadband connection can fail before the mailbox is checked. Receiving servers also consult reputation data for the source IP, so a previously abused address can be refused even when the probe is well formed.

3. Port 25 restrictions. Many cloud providers restrict outbound port 25 by default, especially on new accounts, to limit spam. If your verifier runs there, probes can time out at the network layer. That looks like a mailbox problem but is an infrastructure one. Check your provider's current policy, since these rules change.

4. Missing or mismatched identity. Many servers check that the connecting IP's reverse DNS (PTR) record matches the hostname announced in EHLO. Some also refuse a MAIL FROM domain with no SPF record. A connection with no coherent identity looks like a bot.

5. Greylisting and rate limits. Greylisting asks an unfamiliar sender to retry later, on the basis that legitimate mail servers retry and many spam engines do not. Rate limits do similar work on volume. Both return temporary 4xx replies, which are not verdicts on the address.

Use this table to match symptoms to causes:

Symptom Most likely cause First check
Every probe times out from one server Port 25 blocked or IP listed Test outbound port 25; check the IP on blocklists
Many 4xx replies clustered on one domain Greylisting or rate limit Lower concurrency for that domain; retry later
5.7.x replies across many domains Source IP or identity flagged Check PTR, EHLO match and SPF on your sending domain
Random test addresses accepted Catch-all domain Run catch-all detection before trusting any 250s
Mixed 550 5.1.1 and 250 on one domain Ordinary dead addresses Trust the results; treat as normal list decay

Bottom line: Most blocks come from how your connection looks to the server (identity, reputation, volume), not from the address, so fix the connection first.

Which mail servers block verification probes most?

Blocking is heaviest at large consumer providers under volume, at corporate domains that accept every recipient, and at enterprise gateways that delay their rejections. No provider publishes a probe policy, so treat the patterns below as typical behaviour that practitioners describe, not a rulebook.

Server type Typical response to probes What it means for your results
Large consumer providers (Gmail, Yahoo/AOL, Outlook.com) Throttling and deferrals under volume; clear 550 5.1.1 for many nonexistent mailboxes Throttle replies are not verdicts; re-check later at lower volume. See Yahoo and AOL verification
Microsoft 365 and Exchange tenants Depends on each tenant's configuration; unknown recipients may be refused at RCPT TO or filtered after acceptance Acceptance can be unreliable; run catch-all detection
Self-hosted servers using greylisting First contact returns 4xx, then acceptance on retry A retry after a delay usually resolves these
Corporate catch-all domains Accept any address at SMTP A 250 proves nothing; treat as Risky
Enterprise security gateways Connection-level filtering and late rejection Timeouts or late refusals are common; retry and cross-check

The practical lesson is that the same 250 means different things on different servers. A verifier that reads every acceptance as "valid" is most wrong on the domains where accuracy matters most.

Bottom line: Know the server type before trusting its reply, because a 250 from a catch-all domain is noise.

How do you tell a block from a bad address?

Work through each reply in order. A clean mailbox verdict is final, a catch-all result is downgraded, a connection signal triggers a pause, and anything unresolved waits in Risky.

  1. Check for a clean mailbox verdict. A 250 or a 550 5.1.1 is a usable answer. Record it.
  2. Test whether the domain is catch-all. Probe one random, obviously nonexistent address at the same domain. If it also gets a 250, every 250 on that domain is unreliable, so route those addresses to Risky. The catch-all detection method explains why this single test works.
  3. Treat 5.7.x and 4.7.x replies as connection signals. Stop, check your identity and reputation, and lower volume before retrying.
  4. Retry temporary failures after 30 minutes. Greylisting windows vary by implementation, so no single delay is guaranteed. A 30-minute first retry is a practical default that works across most setups, but it is not a published standard.
  5. Re-check stragglers after 24 hours. Anything still unresolved after two attempts goes to Risky. Do not mark it Invalid, because you have not shown the mailbox is missing.

The rule behind all five steps is simple: only a mailbox-specific answer can move an address to Valid or Invalid. Everything else stays in Risky until you get a clean answer. The RCPT TO explainer and the guide on how an email verifier checks whether an address exists cover the handshake in more depth.

Bottom line: Retry the temporary failures, keep the unresolved ones in Risky, and never treat silence as proof that a mailbox is missing.

How do you reduce blocked probes without evading servers?

The legitimate fixes all make your probing look like a responsible, identifiable sender. None of them hide who you are.

Do not rotate IPs or spread one list across many sources to dodge limits. M3AAWG, the industry group for messaging abuse, describes that kind of distributed evasion as abusive, and it turns a temporary block into a lasting reputation problem.

The architecture decision matters as well. The open-source engine on GitHub lets you run checks yourself, giving you full control and full responsibility for your IP reputation. A hosted service moves probing off your network, but it cannot remove the server's defences, so blocked results still need interpreting.

Approach Strength Specific trade-off
Self-run probes from your own server Full control over timing, logs and retries Your IP reputation and port 25 access determine every result
Open-source engine on your own infrastructure No licence fee for the engine; code you can inspect You own rate limiting, retry scheduling and blocklist monitoring
Hosted verification (Unlimited Verifier) CSV or Excel upload, or the API; Valid/Risky/Invalid output; exports as CSV, Excel or JSON; unlimited standard verification on every plan Blocks still need interpreting; results are kept 24 hours on Free and 30 days on Pro; Pro's 25,000 deep verifications a month must cover any retries

Bottom line: Make your identity, reputation and pace look like a responsible sender, and never disguise the source.

How do you diagnose blocked probes on your own list?

Export the list, verify it, sort the results by domain, and treat Risky rows on problem domains as a diagnostic signal rather than a final verdict. The steps below assume a beginner using a hosted tool.

  1. Export your list as CSV. Keep the email column plus any signup source, so you can trace a problem domain back to the import that brought it in.
  2. Sign up for a free account. The Free plan allows 25 lists a month of up to 150 emails each, with 750 deep verifications and no API access. Results are kept for 24 hours, so export them promptly.
  3. Upload and verify the list. CSV and Excel files up to 10MB are accepted, and the email column is detected automatically. There are no native ESP, CRM or Zapier integrations, so the loop is export, verify, re-import, or calling the API.
  4. Sort the results by domain. Count Valid, Risky and Invalid for each domain. A domain whose Risky share is far above the list average is a candidate for a block or catch-all problem.
  5. Spot-check a suspicious address. Run one address through the free email checker to see how it is classified individually.
  6. Test any catch-all domain. For domains with many Risky results, use catch-all email verification to confirm whether acceptances mean anything.
  7. Re-run the Risky addresses after 24 hours. Count how many move to Valid or Invalid, and keep the rest in Risky. On the Free plan, export before the 24-hour retention window closes.
  8. Automate retries at volume. The email verification API accepts single checks and bulk requests of up to 10,000 emails, and Pro includes 100,000 API emails a month. That suits scripted retry loops within broader API workflows.

Bottom line: Sort by domain, test catch-all behaviour, and re-run unresolved addresses before drawing any conclusion about list quality.

What does a 60,000-row list look like after handling blocked probes?

The figures below are illustrative, not measured results. The plan limits and arithmetic are real. A B2B team has 60,000 addresses, and two large corporate domains produce an unexpectedly high number of Risky results.

Plan fit. The Free plan cannot handle this job. Its 25 lists of 150 emails each cover 3,750 addresses, which is 6.25% of 60,000 (25 × 150 = 3,750; 3,750 ÷ 60,000 = 6.25%). On Pro ($29/month), the file splits into three uploads of 25,000, 25,000 and 10,000 rows, each within the per-list cap.

First pass (standard verification):

Breakdown of the 7,000 Risky results:

Action. The 4,100 catch-all addresses stay in Risky. The 2,900 throttled addresses are re-verified 24 hours later at lower concurrency using deep verification. That uses 2,900 of the 25,000 monthly deep allowance, or 11.6% (2,900 ÷ 25,000).

Second pass. 2,350 of the 2,900 resolve: 1,900 become Valid and 450 become Invalid. The remaining 550 stay in Risky (2,900 − 2,350 = 550).

Final position:

Cost and capacity:

The team ends with 51,100 Valid addresses and a documented reason for the 4,650 it could not confirm. That position is more defensible than either deleting every Risky address or mailing them all. For teams running email verification for ecommerce and SaaS programmes, the same split applies to every large import.

Bottom line: Splitting the work, treating catch-all and throttled results differently, and retrying on a schedule turns a messy first pass into a list you can defend.

What mistakes corrupt probe results?

The most damaging mistakes treat a block as a verdict, so the fixes below all keep "unknown" separate from "invalid."

Bottom line: Most bad results come from treating blocked probes as verdicts, so keep "unknown" separate from "invalid" at every stage.

How we compared (and limitations)

This article is based on public documentation: RFC 5321 for SMTP, RFC 3463 for enhanced status codes, Google's email sender guidelines for authentication and complaint expectations, Spamhaus for blocklist policy, and M3AAWG for abuse norms. We have not run controlled probes against any provider's servers, and server behaviour changes without notice. Recommended retry windows are practitioner defaults, not published standards.

Unlimited Verifier's 99.5% deep verification accuracy is the company's own stated figure, not an independent benchmark. The worked example is illustrative, and the plan limits and per-1,000 costs are computed from list prices.

For the broader picture, read how email verification works, what deliverable means in email verification, deep versus standard verification and send-based catch-all verification. To keep lists healthy over time, see email list maintenance best practices, and for checks at the point of entry, see real-time validation at signup. If you are comparing tools, how to choose an email verifier and best email verification tools are the next steps. Teams with compliance and hygiene obligations should also review email verification pricing before committing to a volume.

How to classify a probe reply

Reply Classification Action
250 2.1.5 Valid (unless catch-all) Keep, flag catch-all domains
550 5.1.1 Invalid Remove from sending list
550 5.7.x Unknown (connection issue) Check IP, DNS, EHLO; retest
421 4.7.0 Unknown (throttled) Retry after 30 minutes
Timeout Unknown Re-check after 24 hours

Frequently asked questions

What is a directory harvest attack?

A directory harvest attack is when a sender guesses large numbers of email addresses at one domain, often using RCPT TO, to find valid mailboxes. Mail servers defend against it by rate-limiting, delaying, or refusing connections that look like guessing.

Why does my verifier get a 550 5.7.1 reply?

A 550 5.7.1 reply is a policy refusal, so it describes your connection or sending behaviour rather than the mailbox. Common causes include a listed IP, missing reverse DNS, or a mismatched EHLO name. Treat it as unknown until retested.

Should blocked verification results be marked invalid?

No. Blocked results should be marked unknown or risky. Only a clean 250 reply or a mailbox-specific 550 5.1.1 reply indicates anything about the address itself.

What does a 421 4.7.0 reply mean?

A 421 4.7.0 reply is a temporary refusal, usually throttling or a sign the server wants you to slow down. Wait and retry later, and do not classify the address as invalid.

Why is port 25 blocked for email verification?

Many cloud hosts and ISPs block outbound TCP port 25 to limit spam, so a verifier on those networks cannot reach the receiving server. The fix is to run probes from a network that allows port 25 or to use a provider that does.

What is greylisting in email verification?

Greylisting makes a server reject first-time connections with a temporary 4xx reply, expecting legitimate senders to retry. Retrying after a delay, commonly around 30 minutes, usually gets a clean answer.

Does catch-all configuration affect verification results?

Yes. A catch-all domain accepts any recipient with a 250 reply, so a positive result does not prove the mailbox exists. Verifiers usually flag these domains as risky instead of valid.

How can I reduce blocking when verifying email lists?

Use fewer concurrent connections per domain, space out probes, keep reverse DNS and EHLO names consistent, and re-check unresolved addresses after a delay. Avoid bulk guessing against one domain.

Does RFC 5321 require servers to answer VRFY?

No. RFC 5321 allows servers to disable VRFY, and many do for privacy and anti-harvesting reasons, so verifiers should rely on RCPT TO responses and treat other outcomes cautiously.

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 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.