Skip to content
  • There are no suggestions because the search field is empty.

What is DMARC and how it works

DMARC takes the SPF and DKIM results a receiver already has and asks one more question: did at least one of them pass for a domain that lines up with the domain in the visible From address? If yes, the message passes DMARC. If neither gives an aligned pass, it fails, and the receiver looks up what you, the domain owner, asked for in your DMARC record: no preference, quarantine, or reject. That's a request, not an order. The receiver makes the final call. If your record asks for reports, receivers that support them send you summaries of who's sending mail as your domain.

For the short definition, see What is DMARC on our main site. This article walks through the mechanics. SPF, DKIM, and how they fit with DMARC are covered in What is email authentication?.

How does DMARC work?

DMARC (Domain-based Message Authentication, Reporting, and Conformance) is defined in the current standard, published in May 2026. That's the revision the working group called DMARCbis. It replaces the older 2015 standard, and reporting now has its own two documents: one for aggregate reports and one for failure reports.

The domain owner publishes one DNS TXT record at _dmarc.<domain>. For example.com that's _dmarc.example.com, and a monitoring record might look like this:

_dmarc.example.com.  TXT  "v=DMARC1; p=none; rua=mailto:dmarc-reports@example.com"

(What each part of the record means is explained in DMARC record explained: syntax, tags and examples. Building the record step by step is covered on our DMARC setup page.)

When a message arrives claiming to be from someone@example.com, a receiver that checks DMARC does roughly this:

  1. Finds the Author Domain. That's the domain in the From header, the address people see. Here, example.com.
  2. Looks up the DMARC record for that domain (or its organizational domain).
  3. Collects the authenticated identifiers. DMARC uses exactly two: - SPF: the envelope sender (MAIL FROM, often shown as Return-Path) domain, but only if SPF passed for it. DMARC ignores SPF's HELO check. - DKIM: the d= domain of any DKIM signature that verifies. One valid, aligned signature is enough.
  4. Checks alignment. Does either passing domain align with example.com? By default (relaxed alignment) a domain aligns if it shares the same organizational domain, so mail.example.com aligns with example.com. Strict alignment needs an exact match. Details are in Identifier alignment for SPF and DKIM and Relaxed vs strict alignment.
  5. Decides pass or fail. A message passes if at least one authenticated identifier aligns with the Author Domain. It fails if there are none, or none align.
  6. Applies its own local policy, informed by your p= value, and records the result for reporting.

The difference between the envelope sender and the From header is the part that trips most people up. It's explained in Envelope From vs Header From.

Why alignment is the whole point

Anyone can publish SPF for a domain they own and get an SPF pass. Anyone can sign a message with DKIM for a domain they own. A spammer can send mail with From: billing@example.com, a bounce address at their-domain.example, and a valid DKIM signature for their-domain.example. SPF passes and DKIM passes, but neither is for your domain. Only an aligned pass shows the use of example.com was authorized, which is why DMARC fails that message.

The same logic catches legitimate mail you forgot about. A typical header can show both checks passing and DMARC still failing, because neither domain aligns with the From domain:

Authentication-Results: spf=pass (sender IP is 198.51.100.10)
 smtp.mailfrom=bounces.vendor.example; dkim=pass (signature was verified)
 header.d=vendor.example; dmarc=fail action=oreject
 header.from=example.com;compauth=fail reason=000

The usual real-world cause is a vendor that sends as you but bounces and signs with its own domain (How to add a third-party sender without breaking DMARC).

DKIM vs DMARC

They aren't alternatives. DKIM signs a message with a domain and lets receivers verify the signature. DMARC doesn't sign anything. It checks whether the DKIM signing domain (or the SPF domain) aligns with your From domain, then applies your policy and sends reports. A valid DKIM signature from your email platform's own domain helps that platform, not your DMARC result. DKIM itself is covered in What is DKIM and how it works, and SPF in What is SPF and how it works.

DMARC policy: none, quarantine, reject

The p= tag states your preference for mail that fails DMARC. There are three values:

  • p=none: the domain owner offers no expression of preference. This is monitoring mode. A discovered p=none policy must not modify existing mail handling.
  • p=quarantine: you consider failing mail suspicious. Receivers may treat it that way, for example by filing it in a junk folder or holding it. What "quarantine" means in practice is up to the receiver (What happens to failing mail under each policy).
  • p=reject: you consider a failure "a clear indication that the use of the domain name is not valid."

Quarantine and reject are enforcement. Other tags, such as sp and np for subdomains, are in DMARC tags explained: p, sp, np, adkim, aspf, rua, ruf, fo, and t and Subdomain policy (sp).

Receivers decide

The final handling of any message is always a matter of local policy. A receiver can deliver mail that fails under p=reject, and it can junk mail that passes. In fact, receivers must not reject messages solely on the basis of a p=reject policy. Your policy is one input to their decision, not the decision.

A pass doesn't mean the mail is good, either. It means only that the use of your domain was authorized. It carries no judgment about whether the message is wanted, so it's no promise of inbox placement.

What reports give you

Reports only arrive if you ask for them, and only from receivers that send them.

  • Aggregate reports (rua): XML summaries from each receiver that sends them, normally daily. Receivers should send them at least once every 24 hours. They list sending IP addresses, message counts, SPF and DKIM results, and alignment. If your record has no rua, receivers must not send aggregate reports at all.
  • Failure reports (ruf): per-message reports about individual failures. They're optional for receivers, and many large-scale providers limit or entirely disable them because they can contain personal data.

The two report types and the authorization record for external report addresses are covered in rua, ruf, and report authorization. What's inside an aggregate report is in Aggregate (RUA) vs forensic (RUF) reports and A field-by-field walk-through of a DMARC aggregate report. Aggregate reports are how you find every service sending as your domain before you tighten the policy. Without rua, p=none tells you nothing.

What DMARC doesn't cover

DMARC protects the exact domain in the From address, and only at receivers that honor an enforcement policy. It doesn't stop:

  • Display-name spoofing. "Example Billing" <alerts@random.example> uses someone else's domain. Display names are out of scope for DMARC.
  • Look-alike domains. Your record covers example.com and its subdomains, not examp1e.example.

A safe starting order

Using both SPF and DKIM is the recommended setup, and domain owners usually start with p=none. In practice:

  1. Make sure SPF and DKIM pass for domains that align with your From address, for every service that sends as you. Wait 48 hours after setting up SPF and DKIM before you add DMARC.
  2. Set up a mailbox or reporting service for aggregate reports.
  3. Publish p=none with rua.
  4. Read the reports, fix the sources that don't align, and only then move toward enforcement. This can take many months. The rollout is covered in A safe rollout: p=none to quarantine to reject.

Two cautions before p=reject. A domain at reject must not rely on SPF alone, so its mail needs aligned DKIM. And domains for general-purpose email should not deploy a DMARC policy of p=reject, because it can break mail for users who post to mailing lists. Reject is a decision to make per domain, after the reports show it's safe. The full sequence is in The right order: SPF, DKIM, DMARC, then BIMI.

If BIMI is on your list, it needs DMARC at enforcement first (p=quarantine or p=reject), not p=none (What is BIMI, and why DMARC comes first).

Misconceptions worth clearing up

  • "DMARC only fails if SPF and DKIM both fail." It fails when neither gives an aligned pass. Both can pass for other domains and DMARC still fails.
  • "p=none protects me." It asks for no change in handling. It's for collecting reports, and staying there indefinitely leaves your domain open (Why staying at p=none is a problem).
  • "p=none on its own gives me data." Not without rua.
  • "My ESP passes SPF and DKIM, so I'm covered." Only if it bounces or signs with your domain, or a subdomain of it.
  • "Quarantine means the spam folder." It means the receiver may treat the mail as suspicious. How is up to them.

Free DMARC account

Turn raw DMARC reports into a readable dashboard

Point your rua address to our platform and read the results by sender and source instead of opening XML files.

  • Results grouped by sending source
  • SPF, DKIM and alignment results per sender
  • The original XML reports are still there to download

Create a free DMARC accountTalk to an expert

FAQ

Does DMARC fail only if both SPF and DKIM fail?

No. It fails if neither produces a pass for a domain aligned with the From domain. Both can pass for unrelated domains and DMARC still fails.

Does p=reject guarantee spoofed mail is blocked?

No. Receivers decide. Some will quarantine or deliver instead.

Will DMARC get my mail into the inbox?

There's no guarantee. A pass only shows the use of your domain was authorized. Where mail lands depends on much more than authentication.

Do I need both SPF and DKIM?

One aligned pass is enough for DMARC to pass. Using both is recommended, and aligned DKIM is required if you publish p=reject.

What is DMARCbis?

The working-group name for the current DMARC standard, published in May 2026, with separate documents for aggregate reports and failure reports. It replaces the older 2015 standard.

Do I get reports automatically?

No. Only if your record includes rua, and only from receivers that send them.