Skip to content
AdminKit

What does an SPF check result actually mean?

An SPF check doesn't return a single pass/fail answer — RFC 7208 defines seven outcomes (Pass, Fail, SoftFail, Neutral, None, PermError, TempError), and which one you get depends on whether the sending IP matches your record's mechanisms and which qualifier is attached to the match. AdminKit's SPF check tool reports exactly which outcome your own record currently produces.

By AdminKit team, Netbase JSC · Updated

Key takeaways

  • RFC 7208 defines seven SPF results. Which one a message gets depends on the connecting IP, the record, and whether evaluation could finish.
  • Pass only covers the envelope sender (MAIL FROM), not the From: address people see. DMARC alignment closes that gap.
  • Fail (-all) and SoftFail (~all) both mean “not authorized”. RFC 7208 says receivers should not reject on SoftFail alone.
  • Neutral must be treated exactly like None, so a record ending in ?all, or with no all at all, protects nothing.
  • PermError means the record is broken for every sender. Too many DNS lookups and duplicate SPF records are the usual causes, and AdminKit’s SPF check reports both.

SPF is a set of outcomes, not a single answer

Unlike a simple yes/no check, RFC 7208 (also tracked in the IETF Datatracker) defines SPF evaluation as producing one of seven named results: Pass, Fail, SoftFail, Neutral, None, PermError and TempError. A receiving server computes one of them for each message, from the connecting IP address, the domain in the SMTP envelope, and that domain’s published record. Knowing which result you’re looking at, and why, matters more than just seeing “it looks fine,” because several of these results look superficially similar but mean very different things to a receiving mail server.

What do the seven SPF results mean?

The definitions below are from RFC 7208 §2.6, with the handling guidance from §8:

ResultRFC 7208 definition (§2.6)Typical causeHandling guidance (§8)
Pass“the client is authorized to inject mail with the given identity”The IP matched a mechanism with + (or no) qualifierFurther policy checks “can now proceed with confidence”
Fail“the client is not authorized to use the domain”The IP fell through to -allLocal policy; reject with 550 5.7.1 if rejecting
SoftFail“probably not authorized”The IP fell through to ~all“SHOULD NOT reject the message based solely on this result”
Neutral“not asserting whether the IP address is authorized”?all, or no all and no redirect“MUST be treated exactly like the ‘none’ result”
NoneNo valid domain, or “no SPF records were retrieved”No v=spf1 record publishedNo information either way
TempError“a transient (generally DNS) error”A lookup timed out or failedAccept, or defer with 451 4.4.3
PermError“published records could not be correctly interpreted”Duplicate records, too many lookups, bad syntaxRequires DNS operator action; 550 5.5.2 if rejecting

RFC 7208 is explicit that “the protocol establishes no normative requirements for handling any particular result.” The handling column is guidance. Each receiver sets its own policy, which is why the same Fail can be rejected by one provider and filed as spam by another.

What does AdminKit’s SPF check actually evaluate?

AdminKit’s SPF check parses your domain’s SPF record with its own code. It doesn’t take a sending IP, so it can’t tell you whether one particular server would get Pass. What it does tell you is the part of the result that comes from the record itself:

What the check reportsWhat it means for the SPF result
No v=spf1 record foundEvery message from this domain gets None
More than one SPF recordEvery message gets PermError (RFC 7208 §4.5)
DNS lookup count above 10Receivers must return PermError once the limit is exceeded (RFC 7208 §4.6.4)
-all, ~all, ?all or +allThe result for any IP not matched earlier: Fail, SoftFail, Neutral or Pass
No all mechanismUnmatched IPs get Neutral, the default in RFC 7208 §4.7

The check follows every include and redirect it finds, counts each include, a, mx, ptr, exists and redirect term against the limit, and lists every mechanism it saw. When a redirect is used, the target record’s all counts, because that is the record whose result applies. It does not count the address lookups an mx term makes for each MX host, or “void lookups” (see below), so a record that passes the check can still fail one of those secondary limits.

Here is an illustrative record using documentation names and addresses, with how the check would count it:

example.com.  TXT  "v=spf1 ip4:192.0.2.0/24 mx include:_spf.provider.example ~all"

ip4:192.0.2.0/24             no DNS lookup
mx                           1 lookup
include:_spf.provider.example  1 lookup, plus every lookup inside that record
~all                         unmatched senders get SoftFail

If _spf.provider.example itself contains three include terms, the total is 5. Hosted providers change their own records, so the count can rise without anyone editing yours.

Pass: what is actually authorized?

A Pass means the connecting server’s IP matched one of the mechanisms in your SPF record — an ip4, ip6, a, mx, include, or exists mechanism resolved to include that address. This is the result every legitimate sending source should produce. It’s worth being precise about what Pass actually verifies: it confirms the IP is authorized to send for the domain used in the SMTP envelope (the “return-path” or “mail from” domain), not the domain shown in the message’s visible From: header a recipient sees — those can differ, which is exactly the gap DMARC’s alignment check exists to close. RFC 9989, the current DMARC specification, says DMARC “relies solely on SPF validation of the MAIL FROM identity”, so an SPF pass on the HELO name alone never counts for DMARC. DMARC alignment: relaxed vs strict explained covers how close that match has to be.

An include only produces Pass when the included record itself returns Pass. RFC 7208 §5.2 notes that “the name ‘include’ was poorly chosen”: a -all inside an included record doesn’t end evaluation of yours, it just means “no match here” and your record carries on to its next term.

Fail and SoftFail: how strong is “not authorized”?

Both results mean the sending IP didn’t match anything in the record, but the SPF record’s own trailing all mechanism decides which one gets returned, via its qualifier. RFC 7208 §4.6.2 maps the - qualifier to Fail (-all), which §2.6.4 defines as “an explicit statement that the client is not authorized to use the domain” — the strongest signal SPF can send. SoftFail (~all) is explicitly described as only “a weak statement,” meant to flag a message as suspicious (commonly routed to spam) without outright rejecting it — useful while a domain owner is still confirming every legitimate sending source is listed, before committing to a hard -all.

RFC 7208 §8.4 shows what a rejection on Fail looks like when the receiver includes the domain’s explanation text. Quoted from the RFC:

550 5.7.1 SPF MAIL FROM check failed:
550 5.7.1 The domain example.com explains:
550 5.7.1 Please see http://www.example.com/mailpolicy.html

For SoftFail, §8.5 says receivers “SHOULD NOT reject the message based solely on this result, but MAY subject the message to closer scrutiny than normal.” The choice between the two is about how sure you are that your record is complete. If you’re not, ~all lets mail from a forgotten sender through while you find it. Microsoft’s SPF setup guidance for Microsoft 365 adds a DMARC angle: DMARC treats both -all and ~all as SPF failures, and Microsoft recommends -all so DMARC can act on messages that fail SPF when they also lack a DKIM signature. For a side-by-side look at when to publish each ending, and how to move from ~all to -all safely, see SPF ~all vs -all: softfail vs hardfail explained.

Neutral and None: why do they get the same treatment?

Neutral results from an explicit ?all qualifier, and is defined as the domain owner stating it has no opinion on whether the connecting host is authorized. None is different: it means no SPF record exists for the domain being checked at all, rather than one that exists but declines to assert anything. RFC 7208 §8.2 removes any practical difference: a Neutral result “MUST be treated exactly like the ‘none’ result; the distinction exists only for informational purposes.”

If a domain’s default all mechanism is left off entirely (no all term at all), evaluation falls through to Neutral behavior by default. RFC 7208 §4.7 calls this an “implicit ‘?all’” and recommends ending the record with an explicit all or redirect because “it aids debugging efforts”. AdminKit’s check flags a record with no all for that reason.

A record with +all is the opposite problem: every IP on the internet gets Pass. The check reports it as almost always a misconfiguration.

What causes a PermError?

A PermError is categorically different from the results above — it means SPF evaluation could not be completed correctly because of a problem with the record’s configuration, not because of anything specific to the sending IP. RFC 7208 names several causes:

CauseRuleDoes AdminKit’s check catch it?
Two or more v=spf1 records at the same name§4.5: the set “includes more than one record”Yes, reported as multiple records
More than 10 lookup-causing terms§4.6.4: “MUST return ‘permerror’”Yes, lookup count shown against the limit
More than two “void lookups” (empty or NXDOMAIN answers)§4.6.4: “SHOULD limit ‘void lookups’ to two”No
An include that points at a domain with no SPF record§5.2: a recursive None makes the include return permerrorNo; the included domain simply adds no terms
Syntax errors in a term§2.6.7: the records “could not be correctly interpreted”Partly; unknown terms are listed but not flagged

The lookup limit is the one most often hit in practice. Google Workspace’s own SPF troubleshooting documentation confirms this second cause in practice, warning that nested lookups inside an include mechanism count toward the same 10-lookup budget and are a common, easy-to-miss way real-world records exceed it. Google’s page also says each domain “can have only one SPF record”. See fixing SPF “too many DNS lookups” (PermError) for how the lookup limit gets exceeded and how to fix it. To test a draft record before you publish it, AdminKit’s SPF validator checks pasted record syntax and counts lookup-causing terms offline, without following includes.

What causes a TempError?

A TempError means something transient went wrong during evaluation — typically a DNS query that timed out or returned a server failure — rather than anything wrong with the record’s actual content. RFC 7208 §8.6 says the receiver “can choose to accept or temporarily reject the message”, and if it rejects, it should use SMTP code 451 with enhanced status code 4.4.3. A 4xx code tells the sending server to queue and retry, so a TempError usually delays mail instead of losing it.

RFC 7208 §4.6.4 also suggests receivers put a time limit on the whole evaluation, of at least 20 seconds, and return TempError if it runs over. A record with long include chains that resolve slowly is more exposed to that limit. A TempError that appears once and doesn’t recur usually isn’t worth investigating; one that persists across repeated checks points to a DNS resolution problem independent of the SPF record’s syntax.

Where do you see the result a receiver actually got?

The receiver records its verdict in the message headers, so the authoritative answer for one message is in that message. RFC 7208 §9 defines a Received-SPF header and also shows the shorter Authentication-Results form from RFC 8601. Both examples below are quoted from RFC 7208 §9:

Received-SPF: pass (mybox.example.org: domain of
 myname@example.com designates 192.0.2.1 as permitted sender)
    receiver=mybox.example.org; client-ip=192.0.2.1;
    envelope-from="myname@example.com"; helo=foo.example.com;

Authentication-Results: myhost.example.org; spf=pass
  smtp.mailfrom=example.net

smtp.mailfrom and envelope-from name the domain whose record was evaluated, which isn’t necessarily your From: domain. A hosted sending service can use its own bounce domain there, in which case the SPF result is about the provider’s record, not yours. Paste the full header block into AdminKit’s email header analyzer to see the SPF, DKIM and DMARC verdicts the receiving servers recorded, parsed in your browser.

Reading your own result

Run AdminKit’s SPF check against your domain to see which of these seven outcomes your current record produces for senders it doesn’t list, its lookup count, and whether it would cause None or PermError for everyone. If the record ends in -all (or ~all while you’re still finding senders) and stays under the limit, SPF is doing its job; confirm each legitimate sender by checking for spf=pass in a message it sent. If it’s PermError, fix the underlying record problem before anything else — a broken record can make even your legitimate mail fail. For how SPF fits together with DKIM and DMARC as a complete authentication setup, see SPF, DKIM and DMARC explained. This guide is part of AdminKit’s email authentication cluster of SPF, DKIM and DMARC guides and tools.

This guide is general information, not a security audit of your specific network — see our disclaimer for what AdminKit's guidance does and does not commit to.

Frequently asked questions

What's the actual difference between Fail and SoftFail?

Both mean the sending IP isn't authorized, but they carry different instructions to the receiver. RFC 7208 §2.6.4 defines Fail (-all) as "an explicit statement that the client is not authorized," which receivers are expected to treat as a real rejection signal, while SoftFail (~all) is "a weak statement" meant to be accepted but marked, typically routed to spam rather than bounced outright.

My check came back Neutral — does that mean something is wrong?

Not necessarily broken, but it means SPF isn't doing much. Neutral (explicit ?all, or the default behavior when a record has no all mechanism at all) tells receivers the domain is explicitly not asserting anything about whether the sender is authorized — mail is treated the same as if no SPF record existed. It's a valid configuration, but it provides none of SPF's actual protection.

What causes a PermError, and is it worse than a Fail?

In one important way, yes: PermError isn't about one sending IP being unauthorized, it means the record itself is broken: more than one SPF TXT record published at the same name, unparseable syntax, or more than 10 DNS lookups during evaluation (RFC 7208 §4.6.4). It affects every sender, authorized or not, because the check can't be completed for anyone. RFC 7208 §8.7 says a receiver that rejects for this reason should use SMTP code 550, the same code it uses for Fail, and for DMARC a PermError is simply not an SPF pass.

What does a TempError mean, and should I worry about it?

TempError means evaluation hit a temporary problem — usually a DNS lookup that timed out or returned a server failure — not a problem with the record's content. RFC 7208 §8.6 lets the receiver either accept the message or temporarily reject it with SMTP code 451, which tells the sending server to queue it and try again later. A one-off TempError usually isn't worth chasing; a persistent one suggests a DNS resolution problem worth investigating separately.

Does getting a Pass result mean the email definitely isn't spoofed?

No. A Pass only confirms the connecting server's IP is listed as authorized in the SPF record for whatever domain was checked during the SMTP transaction (the envelope-from domain) — it says nothing about whether that domain matches the visible From: address the recipient actually sees, and nothing about whether the message content was altered. Those two gaps are exactly what DMARC alignment and DKIM close; see SPF, DKIM and DMARC explained for how the three combine.

Can AdminKit tell me what result a specific server's IP will get?

No. AdminKit's SPF check reads your published record and follows its include and redirect chain, so it tells you what unlisted senders get (from the all qualifier) and whether the record itself produces None or PermError. It doesn't take a sending IP as input, so it can't say whether one particular server would get Pass. For that, look at the Received-SPF or Authentication-Results header of a message that server actually sent.

Want AdminKit to watch this for you?

Free checks tell you the state today. Protective DNS and monitoring tell you when it changes.

Huy Nguyen (David), Founder and CEO of Netbase JSC

Talk to an expert

Huy Nguyen (David), Founder & CEO, Netbase JSC

Ask whether this fix applies to your setup, or get a second pair of eyes on your DNS and mail records.