How do you protect a parked domain from email spoofing?
Publish four DNS records: a null MX (MX 0 .), an SPF record of v=spf1 -all, a DMARC record with p=reject and, if you follow UK government guidance, an empty DKIM key. Together they tell receivers the domain sends and accepts no mail. AdminKit's free MX, SPF and DMARC checks confirm each record.
By AdminKit team, Netbase JSC · Updated
Key takeaways
- A domain that never sends or receives mail should say so in DNS. Silence lets anyone forge it, and receivers have nothing to check the forgery against.
- The core set is three records: a null MX, an SPF record of
v=spf1 -alland a DMARC record withp=reject. GOV.UK adds a fourth, an empty DKIM key; M3AAWG says to publish no DKIM key at all. - SPF and MX records do not cover subdomains. Use wildcard records where your DNS host supports them, and list existing subdomains by name.
- After publishing, check each record with AdminKit’s free MX lookup, SPF check and DMARC check.
What counts as a parked domain?
A parked domain is any domain you own but do not use for email. M3AAWG’s Protecting Parked Domains Best Common Practices gives the typical cases: a defensive registration bought so that nobody else can abuse the name, a domain registered for later use, and a domain or subdomain that only hosts a website. The UK government’s guidance on protecting domains that do not send email adds a fourth case that is easy to forget: a legacy domain that used to send mail and no longer does, for example after a rebrand or a move to a new primary domain.
What these domains have in common is that no legitimate message should ever carry them in the From address or the envelope sender. That makes them easy to protect. A domain that does send mail needs a careful list of every service that sends for it. A parked domain needs a list with nothing on it.
Why is a domain that sends no mail a spoofing risk?
Receivers decide whether a message really comes from a domain by looking up that domain’s SPF, DKIM and DMARC records. If the domain publishes none of them, there is no policy to fail. A forged message is judged only on content and reputation, and a clean-looking look-alike domain can do well on both.
M3AAWG’s problem statement is direct: failing to publish DNS records that signal a parked domain does not send or receive email “provides opportunities for third parties to abuse domains of this nature”, and domains similar to well-known domains are especially vulnerable. That is exactly the kind of domain brand owners park on purpose: misspellings, other top-level domains and old product names. The typosquatting registrations you bought to keep away from attackers can still be used in a forged From address if you leave them silent.
Which DNS records does a parked domain need?
Both government and industry guidance land on the same core set. The table compares the GOV.UK and M3AAWG recommendations record by record.
| Record | Name | Value | GOV.UK | M3AAWG |
|---|---|---|---|---|
| Null MX | the domain itself | 0 . | Yes, if your DNS provider supports it | Yes, plus a wildcard |
| SPF (TXT) | the domain itself | v=spf1 -all | Yes | Yes, plus a wildcard |
| DMARC (TXT) | _dmarc | v=DMARC1; p=reject plus optional tags | Yes, with sp=reject and strict alignment | Yes, rua optional |
| Empty DKIM key (TXT) | *._domainkey | v=DKIM1; p= | Yes, and revoke existing selectors | No, publish no DKIM record |
Here is a complete zone fragment for a parked domain, using the reserved example name:
example.com. 3600 IN MX 0 .
example.com. 3600 IN TXT "v=spf1 -all"
*.example.com. 3600 IN MX 0 .
*.example.com. 3600 IN TXT "v=spf1 -all"
_dmarc.example.com. 3600 IN TXT "v=DMARC1; p=reject"
Each line does one job, and the sections below explain them in turn. Most DNS hosting panels ask for the same three things: the record type, the host or name (often @ for the domain itself) and the value.
What does a null MX record do?
RFC 7505 defines the null MX: a single MX record with preference 0 and the root name . as the mail exchanger. Because . is not a valid host name, the record cannot be mistaken for a real mail server. The RFC adds that a domain publishing a null MX must not publish any other MX record.
The null MX covers the receiving side. Without any MX record, a sender falls back to the domain’s A or AAAA address and tries to deliver there, which for a website-only domain means queued retries against a web server. With a null MX, senders know at once that the address cannot receive mail. RFC 7505 §4.1 also covers the other direction: it gives the reply code a server should use when it rejects incoming mail because the envelope sender or From domain has a null MX, and §4.2 says a domain that publishes a null MX and still sends mail risks having that mail rejected. That is why a null MX is only for domains that never send. Never add one to a domain you use in From addresses.
GOV.UK notes that some DNS providers do not support a null MX record and says not to worry if you cannot create one. The SPF and DMARC records still do the anti-spoofing work. For more on how MX records and their preference values work, see what MX record priority actually does.
Why “v=spf1 -all” and not “~all”?
An SPF record lists the servers allowed to send as a domain and ends with an all term that catches every server not listed. v=spf1 -all has an empty list and a hardfail ending, so every server on the internet gets the SPF result Fail. M3AAWG calls this a “naked” -all, and RFC 7505 §4.2 points operators of domains that do not send mail to the same SPF policy.
On an active domain, many administrators keep ~all while they confirm every sender, because a forgotten service would otherwise fail. A parked domain has no senders to forget, so there is no reason for the softer ending. SPF ~all vs -all explains the difference in detail, and its decision table gives -all for a domain that never sends mail.
Microsoft’s SPF setup guidance includes a “parked domains” scenario with exactly this record for each unused domain. It also points out that each subdomain needs its own SPF record, and that for a parked domain “it’s virtually impossible to guess which subdomains might be needed”, so it suggests a wildcard record if your DNS host supports one.
What should the DMARC record say?
SPF alone checks only the envelope sender, not the From address people read. DMARC closes that gap: it passes only when SPF or DKIM passes for a domain aligned with the visible From domain. On a parked domain with -all and no usable DKIM key, nothing can pass. M3AAWG puts it this way: it “would be impossible to get an SPF or DKIM pass on an aligned domain”, so forged mail fails DMARC, and p=reject asks receivers to reject it.
The minimal record is v=DMARC1; p=reject. Microsoft’s DMARC setup guidance recommends exactly that for registered domains that no one should expect mail from. GOV.UK’s version for government domains adds more:
v=DMARC1;p=reject;sp=reject;adkim=s;aspf=s;fo=1;rua=mailto:dmarc@yourdomain.example
The sp=reject tag sets the policy for existing subdomains, and adkim=s and aspf=s require strict alignment. Under RFC 9989, the current DMARC specification, the p policy already applies to subdomains when sp and np are absent, so p=reject alone covers them. The np tag sets a policy for subdomains that do not exist at all, and also falls back to sp or p.
The rua reporting address is optional. M3AAWG explains that aggregate reports are mainly for tuning real mail streams, which a parked domain does not have, though some owners like to see how much forged mail receivers are rejecting. If you add one, it has to point at a domain that receives mail. When that is a different domain, the receiving domain must publish an authorization record, which M3AAWG shows as:
example.com._report._dmarc.example.net. TXT "v=DMARC1;"
DMARC aggregate reports explains what the reports contain.
Should you publish an empty DKIM key?
This is the one point where the main sources disagree, and both positions are reasonable.
GOV.UK says to create a wildcard TXT record at *._domainkey with the value v=DKIM1; p= and to revoke all existing DKIM selectors. An empty p= tag is defined in RFC 6376 as a revoked key, so any signature that claims a selector under the domain fails verification. GOV.UK says this makes receivers more likely to reject mail from the domain.
M3AAWG takes the opposite view: there is no need to publish any DKIM record for a parked domain, and none should be published. If no key exists, a forged DKIM signature fails anyway, because the receiver cannot find a public key to check it with. Microsoft adds a related point: DKIM CNAME records are not recommended for parked domains.
In practice the outcome is the same. A forged signature cannot verify, so DKIM cannot rescue a forged message from your DMARC p=reject. The step both sources agree on matters more: if the domain used to send mail, remove or revoke the old selectors so a leftover key cannot be reused. AdminKit’s DKIM check looks up a given selector and reports an empty p= as a revoked key.
How do you cover subdomains?
DMARC finds the organizational domain’s record and applies it to subdomains, so _dmarc.example.com covers mail.example.com unless that subdomain publishes its own DMARC record. SPF and MX do not work that way. A receiver checks the SPF record of the exact domain in the envelope sender, so invoices.example.com with no SPF record has no SPF policy at all.
Wildcard records fill the gap. M3AAWG’s examples publish *.example.com with both MX 0 . and v=spf1 -all. There is one catch: a wildcard DNS record only answers for names that have no records of their own. If www.example.com has an A record, the wildcard does not supply MX or TXT for it. That is why M3AAWG’s example for a domain with a website lists the existing subdomain separately:
sub.example.com. IN A 192.0.2.10
sub.example.com. IN MX 0 .
sub.example.com. IN TXT "v=spf1 -all"
The address is from the documentation range. Repeat the MX and SPF lines for every subdomain that already exists in the zone.
How do you protect many parked domains at once?
Brand owners often hold dozens of defensive registrations. M3AAWG’s example for multiple parked domains points each domain’s _dmarc name at one shared record with a CNAME:
_dmarc.example.com. IN CNAME _dmarc.parked.example.net.
_dmarc.example.org. IN CNAME _dmarc.parked.example.net.
_dmarc.parked.example.net. IN TXT "v=DMARC1; p=reject"
One record then sets the policy for all of them, and a change of reporting address happens in one place. The null MX and SPF records still go on each domain, because they describe that domain itself. GOV.UK makes the same practical point from the other direction: you can make the changes in bulk and alter them later if you decide to start sending mail from one of the domains.
M3AAWG also suggests putting a contact address for the domain in its SOA record, because role addresses such as postmaster@ cannot exist on a domain that accepts no mail.
What can these records not do?
These records tell receivers what you want. They do not force anyone to follow it. M3AAWG’s conclusion says there is no guarantee that mailbox providers will make best use of this information. A receiver that does not check DMARC, or that quarantines instead of rejecting, may still deliver a forged message somewhere.
The records also protect only the exact domain you publish them on. They do nothing about a different look-alike domain that someone else registers, and nothing about display names that imitate your brand while using an unrelated address. Monitoring new registrations similar to your name is a separate job.
How do you check a parked domain with AdminKit?
Three free checks, no account needed, cover the core records:
- MX lookup. It reports a null MX explicitly as the domain’s statement that it accepts no mail, rather than as a missing or broken record.
- SPF check. It reads the published record the way a receiver does and explains how strict the trailing
allmechanism is. For a parked domain you want to see-alland no other terms. - DMARC check. It parses the
_dmarcrecord, explains whatp,spandnpdo, and flags an external report address that has not authorized your domain.
To draft the records before you publish them, the SPF generator and DMARC generator build the record strings in your browser. You still publish them at your DNS host and confirm them with the live checks. For how the three standards fit together on a domain that does send mail, read SPF, DKIM and DMARC explained. This guide is part of AdminKit’s email authentication cluster.
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
Why would anyone spoof a domain that never sends email?
Because nothing stops them. A domain with no SPF or DMARC record gives receivers no way to tell forged mail from real mail, so a look-alike or defensively registered domain is a convenient From address for phishing. M3AAWG's parked-domain best practice says failing to publish these records gives third parties opportunities to abuse such domains, especially names similar to well-known brands.
Is a null MX record enough on its own?
No. A null MX (RFC 7505) only says the domain accepts no mail. It says nothing about who may send mail as the domain. Pair it with v=spf1 -all, which authorizes no sending server, and a DMARC record with p=reject, which asks receivers to reject anything that fails.
Should I publish an empty DKIM record or none at all?
The sources differ. GOV.UK guidance for government domains says to create a wildcard empty DKIM key (v=DKIM1; p=) and revoke existing selectors. M3AAWG says no DKIM record is needed for a parked domain and none should be published, since a forged signature fails anyway when there is no key. Either way, a forged DKIM signature cannot verify.
Do these records protect subdomains too?
Partly. A DMARC record at the domain applies to its subdomains unless an sp or np tag says otherwise. SPF and MX do not inherit: each name needs its own record, which is why M3AAWG and Microsoft suggest wildcard records. A wildcard only answers for names that have no records of their own, so list any existing subdomain explicitly.
Do I need a rua address in the DMARC record of a parked domain?
It is optional. M3AAWG calls it optional because a parked domain has no real mail streams to tune, though reports show how much forged mail receivers see. The report address must be at a domain that receives mail, and if it is a different domain, that domain must publish an authorization record for your domain.
How do I check that the records are published correctly?
Run AdminKit's free MX lookup, which reports a null MX explicitly, the SPF check, which explains how strict the all mechanism is, and the DMARC check, which explains the p, sp and np tags and flags report addresses that have not authorized your domain.
Sources
- RFC 7505 — A "Null MX" No Service Resource Record for Domains That Accept No Mail, §3 and §4
- RFC 9989 — Domain-Based Message Authentication, Reporting, and Conformance (DMARC), §4.7 (p, sp, np tags)
- GOV.UK — Protect domains that do not send email (Government Digital Service and CDDO)
- M3AAWG — Protecting Parked Domains Best Common Practices (updated June 2022)
- Microsoft Learn — Set up SPF to identify valid email sources (Scenario: Parked domains)
- Microsoft Learn — Set up DMARC to validate the From address domain (parked domains)
Want AdminKit to watch this for you?
Free checks tell you the state today. Protective DNS and monitoring tell you when it changes.
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.