Skip to main content

Free DMARC Checker and Policy Simulator

Paste a DMARC record to check it against RFC 9989, see the policy it requests for sample SPF and DKIM results, and copy a corrected record. It runs in your browser.

Last updated 15 November 2026

Loading tool…

This DMARC checker reads a DMARC record you paste and checks it against RFC 9989, the Standards Track DMARC specification dated May 2026, which obsoletes RFC 7489. Paste the record with or without its quotes: an answer copied from dig or nslookup works. Each line counts as one record, and a record's quoted strings are joined in order. Each tag gets its own line, and an unknown tag, a repeated tag, a value the specification does not allow and a missing v or p tag each get a finding of their own. A policy simulator then shows the policy the record requests for mail that fails DMARC, for the organisational domain and for subdomains that exist and do not, and for a sample message whose SPF and DKIM results you set. Last, the tool drafts a corrected record from your choices and explains every difference from the one you pasted. It makes no DNS lookup, so it shows the commands to copy the record from DNS yourself.

How to use this tool

  1. 1

    Copy the record from DNS

    Run the dig or nslookup command shown in the tool with your domain in place of example.com, or copy the TXT value at _dmarc and your domain from your DNS host. The tool does not make the lookup.

  2. 2

    Paste it and check the record

    Paste one record, up to 2,000 characters, with or without its quotes, and press Check record. The tool states what RFC 9989 says about the record as a whole, and marks each finding to fix, to review, a note or a match, with the reason.

  3. 3

    Try sample messages in the simulator

    A table shows the policy the record requests for each kind of From domain. Then set the From domain, the SPF and DKIM results and how each checked domain relates to the From domain, and read what the record requests.

  4. 4

    Draft and copy a corrected record

    Choose the policies, test mode, alignment and up to five aggregate report addresses. The tool writes the record, lists each difference from the one you pasted, and gives you a copy button.

What the DMARC record checker reads, and how

A DMARC record is a TXT record in DNS. RFC 9989 says the record for example.com is published at _dmarc.example.com, and that when a TXT record holds several character-strings, the module evaluating DMARC joins them in order and parses the result as one string. The checker does the same within each record, reading each line of a paste as one record and lines inside parentheses as one line, so an answer copied from dig, nslookup or a zone file can keep its quotes and the record name in front of them. From a paste of several lines it keeps the one that starts with v=DMARC1, since section 4.10 discards records that do not, and it refuses a paste where more than one does, since that section then discards them all. It then splits the record into tag=value pairs at each semicolon, as the grammar in the specification does. The version tag comes first. RFC 9989 says v is required, must be the first tag and has one possible value, DMARC1, which is case sensitive; if any of that is wrong, a receiver must ignore the entire record. Every other tag gets a line saying what it does. Five kinds of problem get findings of their own. An unknown tag must be ignored, and RFC 9989 says syntax errors are discarded in favour of a tag's default value, if it has one, or ignored. A repeated tag is more serious: RFC 9989 says the record follows the DKIM tag-value syntax, and RFC 6376 says that when a tag name occurs more than once, the entire tag-list is invalid. A bad value, such as p=strict, is checked against the values the specification lists. A value in another letter case, such as p=Reject, is read in lower case and flagged for review: RFC 9989's grammar marks only DMARC1 as case sensitive, while RFC 6376 treats values as case sensitive unless a tag says otherwise, so write values in lower case. The fifth is a missing v or p tag, covered below.

RFC 9989 replaced RFC 7489, and three tags went with it

RFC 7489, dated March 2015, is an Informational document published as an independent submission. RFC 9989, dated May 2026, is an Internet Standards Track document, and its header says it obsoletes RFC 7489. Two companion documents, RFC 9990 for aggregate reports and RFC 9991 for failure reports, say they obsolete RFC 7489 as well. The checker follows RFC 9989, and where a finding rests on another document, such as RFC 7489, RFC 6376 or RFC 9990, it names that document. Three tags are gone. pct asked receivers to apply the policy to a percentage of failing mail, rf named the failure report format and ri asked for an interval between aggregate reports. RFC 9989 removed all three, and the DMARC Tags registry lists them as historic, meaning deprecated and not expected to be in use in any current implementation. RFC 9989 explains that operational experience showed pct was usually not accurately applied unless its value was 0 or 100. In its place is t, a test mode. With t=y, the Domain Owner asks the receiver not to apply the policy but any special handling rules it has, such as rewriting the From header field, and expects failing mail to get the policy one level below the one written: quarantine becomes none and reject becomes quarantine. RFC 9989 says t=y and t=n are meant to be analogous to pct=0 and pct=100. Two more tags are new: np, the policy for subdomains that do not exist, and psd, which marks a public suffix domain. p changed as well. RFC 7489 called it REQUIRED for policy records; RFC 9989 calls it RECOMMENDED, and says that when p is missing or invalid, or sp or np is invalid, a receiver acts as if the record held p=none when rua has a syntactically valid address, and applies no DMARC processing when it does not. When your record still has pct and asks for quarantine or reject, the simulator also shows how RFC 7489 told receivers to use it.

How a record turns SPF and DKIM results into a policy

DMARC builds on two existing checks, SPF and DKIM, which RFC 9989 says provide domain-level authentication but are not directly associated with the domain in the From header field. DMARC relies solely on SPF's check of the MAIL FROM domain, and on the DKIM signing domain, the d= value in a validated signature. Either one counts only when it passes and its domain is aligned with the From domain. Alignment has two modes. Strict means the two domains are identical. Relaxed, the default, means they have the same organisational domain, so an identifier of foo.example.com aligns with a From domain of news.example.com, the example RFC 9989 gives. adkim sets the mode for DKIM and aspf for SPF. A message passes DMARC when at least one aligned pass exists, and fails when none does. For a failing message, the record names a policy. When the From domain is the domain the record is published for, p applies. For a subdomain with no record of its own, RFC 9989 takes the policy from sp when the subdomain exists and from np when it does not; without np, sp applies to subdomains that do not exist too, and without sp or np, p applies. With t=y, the Domain Owner expects failing mail to get the policy one level lower. As a DMARC policy checker, the simulator applies these rules to a sample message. For SPF it offers pass or one of fail, softfail, neutral and none, four of the results RFC 7208 lists; the two error results are left out. What it reports is what the record requests. RFC 9989 says the final handling of any message is left to the receiver's local policy, that receivers must not reject a message solely because of p=reject, and that without other knowledge and analysis they must treat that failing mail as if the policy were quarantine. It also says receivers may still reject or quarantine a message that passes.

Report addresses, including ones in another organisational domain

rua asks receivers for aggregate reports, and ruf for failure reports about individual messages. RFC 9989 says that without rua, receivers must not generate aggregate reports for the domain, and that a Domain Owner fully participating in DMARC must set up a mailbox to receive aggregate reports and collect and analyse them. Receivers that send reports must support mailto addresses and must ignore a scheme they do not support, so the checker notes any address that is not mailto. It also flags an address that fails its own address test, which is simpler than full URI syntax: the test looks for a scheme, no spaces or other characters it does not accept, a percent sign only before two hexadecimal digits, no unencoded exclamation mark, which RFC 9989 says must be percent-encoded inside an address, and for mailto, one "@" followed by a domain of two or more labels. A trailing size limit such as !10m is obsolete syntax that RFC 9989 says reporters should ignore. Addresses in another organisational domain carry a condition. RFC 9989 says reporting addresses generally have to be inside the domains reports are requested about. Where one is not, RFC 9990 section 4 has the receiver query a TXT record at a name built from the domain the policy came from, then _report._dmarc, then the address's domain. For example.com sending reports to reports@thirdparty.example.net, the name starts with example.com._report._dmarc and ends with thirdparty.example.net. If no record there starts with v=DMARC1, the receiver must ignore the address. RFC 9991 applies the same procedure to ruf. A reporting service willing to take reports for any domain can publish a wildcard record instead, at *._report._dmarc and its own domain. Failure reports need more care than aggregate ones. RFC 9989 notes that actual email content can appear in them, so ruf addresses receive information that might be private, and it says receivers may choose whether to send them at all. The checker reads every address as text and contacts none of them.

What this DMARC checker cannot tell you

The checker works only on the text you paste, in your browser. What you paste and type stays in the page: the tool makes no network request with it and no DNS lookup. That keeps your record on your device, and it also sets the limits. It cannot confirm that the record is published at _dmarc and your domain, or that it is the only DMARC record there. That second point matters, because RFC 9989 says that when more than one DMARC record is returned for a name, they are all discarded. The checker refuses a paste in which more than one record starts with v=DMARC1, but it cannot see a record you did not paste. It cannot tell you what SPF and DKIM return for your real mail, which is why the simulator asks you to choose. It does not look up the confirming record another domain needs before receivers send it your reports, and it does not contact any report address. It cannot say what a particular receiver will do with a message, because RFC 9989 leaves that to the receiver. For a test of a message you actually send, the email spam tester gives you an address to send it to and reports on SPF, DKIM and DMARC among its other checks. The limits on input are deliberate. A paste over 2,000 characters is refused whole rather than checked in part, so nothing is reported for a cut record. The finding list shows at most 40 findings and says when it left some off, and the corrected record takes up to five report addresses in 500 characters. When a check runs, the page sends one analytics event that names the tool and the page, and nothing from your input.

Sources

Worked example

An invented example, not taken from any real domain: a record kept from an older setup guide is checked before its owner relies on the reject policy.

Pasted record
v=DMARC1; p=reject; aspf=s; pct=25; rua=mailto:dmarc@example.com
From domain
example.com, the organisational domain
SPF result
pass, for bounce.example.com
DKIM result
fail

DMARC fail. aspf=s needs the MAIL FROM domain to be identical to the From domain, so the SPF pass does not align, and the record requests reject.

The checker also notes that RFC 9989 removed pct, and the corrected record leaves it out. RFC 9989 says a domain that publishes p=reject must apply valid DKIM signatures and not rely on SPF alone, so the owner's next job is DKIM signing for example.com, before anything else in the record changes.

Who uses this tool

Owners tightening a policy

See what quarantine or reject would request for mail from the main domain and each kind of subdomain before changing p, then copy the new record.

IT providers taking over a domain

Paste a record left by an earlier provider, find tags RFC 9989 removed or does not define, and copy a cleaned-up version with each change explained.

Teams adding a report service

Check that the service's report address reads correctly, and see the name of the confirming record its domain needs before receivers send reports there.

Developers sending from a subdomain

Check the record with adkim=s and again with adkim=r, and simulate mail from a subdomain to see whether a signing domain one level up aligns under each.

Frequently asked questions

Does this DMARC checker do a DMARC lookup for my domain?

No. It reads only the record you paste, in your browser, and makes no DNS lookup. To get the record, run dig _dmarc.example.com TXT +short or nslookup -type=TXT _dmarc.example.com with your domain in place of example.com, or copy the TXT value from your DNS host, then paste the answer. Its quotes and any record name in front of it can stay.

Why does the checker say pct is no longer part of DMARC?

RFC 9989, the Standards Track DMARC specification dated May 2026, removed pct along with rf and ri, and the DMARC Tags registry lists all three as historic. RFC 9989 says operational experience showed pct was usually not accurately applied unless its value was 0 or 100. To trial a stricter policy, it offers t=y, which asks for the policy one level below the one written.

What is the difference between p, sp and np in a DMARC record?

p applies to the domain the record is published for, and to its subdomains unless another tag says otherwise. sp applies to subdomains that exist, and np to subdomains that do not exist in DNS. RFC 9989 says that without np, sp applies to subdomains that do not exist too, and that without sp or np, p applies. The simulator's table shows which tag applies to each kind of From domain.

If my record says p=reject, will receivers reject mail that fails?

Not necessarily. The record states a request. RFC 9989 says the final handling of any message is left to the receiver's local policy, that receivers must not reject a message solely because of p=reject, and that without other knowledge and analysis they must treat that failing mail as if the policy were quarantine. It also says a domain publishing p=reject must apply valid DKIM signatures and not rely on SPF alone.

Why does a DMARC record with a repeated tag get no outcome?

RFC 9989 says DMARC records follow the DKIM tag-value syntax, and RFC 6376, the DKIM specification, says that when a tag name occurs more than once in a tag-list, the entire tag-list is invalid. RFC 9989 itself does not state how a receiver reads such a record, so the simulator gives no outcome for it until the repeat is removed. The corrected record writes each tag once.

Can DMARC reports go to an address at another company?

Yes, with a condition. When the report address is in a different organisational domain from your record, RFC 9990 says the receiver must find a TXT record starting with v=DMARC1 at a name made of your domain, then _report._dmarc, then the address's domain, and must ignore the address without one. The reporting service publishes that record in its own DNS.

Related tools and services