Skip to content
fixGuide

DMARC p=none vs quarantine vs reject

p=none watches without asking receivers to block. Quarantine asks for junk. Reject asks them to refuse the message. Here is when to move, and what still happens at none.

You searched for

“dmarc p=none vs quarantine vs reject”

Last verified: Oct 21, 2026Published: Oct 21, 2026

p= is the only part of a DMARC record most people remember, and the part they change too early. none, quarantine, and reject are instructions to other people’s mail servers. Your monitoring tool does not enforce them.

Check the live record with the DMARC checker before you edit anything. For the report fields that tell you whether a move is safe, see how to read DMARC reports.

What each policy asks for

p= value What receivers are asked to do with DMARC failures What you get
none Do not quarantine or reject because of DMARC Reports, if rua= is set and authorized
quarantine Treat failures as suspicious (typically junk) Reports plus a visible penalty for failures
reject Refuse the message at SMTP Reports plus failures that should not land

Two limits on that table:

  1. Other filters still run at p=none. Gmail’s 550 5.7.26 and bulk-sender rules do not wait for you to leave none. See Fix Gmail error 550 5.7.26.
  2. Receivers can soften policy. pct=, local policy, and ARC on forwards all mean reject is not a laser.

DMARC failures are messages with no aligned SPF or DKIM pass. An unaligned SPF pass still fails DMARC. That is alignment, and it is the usual reason a “working SPF” domain is not ready for quarantine.

A record that only monitors

A starting record looks like:

v=DMARC1; p=none; rua=mailto:[email protected]; adkim=r; aspf=r

Useful additions once you know your tree:

  • sp= if subdomains should not inherit p= blindly
  • pct= later, when you start enforcing
  • fo= only if you also collect forensic (ruf) reports; most domains never need ruf

rua= on a third-party domain needs external destination authorization or many receivers will not send.

When to leave none

Stay at none until all of these are true:

  1. You can name every source with meaningful volume.
  2. Legitimate senders show aligned SPF or DKIM (both is better; DKIM is what usually survives a forward, because SPF is checked at the last hop).
  3. Remaining failures are small or obviously spoofed.
  4. You have watched a real sending week, not three quiet midweek days.
  5. You know how to publish the previous record again if a forgotten CRM appears.

Then publish quarantine (or reject with a low pct=) and watch. If a known tool starts landing in junk, fix its alignment, then keep going. If a random ISP blast disappears, that is the policy doing the job.

A common path:

p=none
→ p=quarantine; pct=10
→ p=quarantine; pct=100
→ p=reject; pct=10
→ p=reject

You can skip steps if the report picture is boring and you accept the blast radius. You cannot skip reading the reports.

What reject does not do

Reject does not stop someone from sending as you. It asks receivers who honor DMARC to refuse unauthenticated copies. Some receivers lag, some ignore pct=, and some will still deliver a forwarded copy with a surviving DKIM signature (that copy passed DMARC).

Reject also does not replace SPF or DKIM. If those records are wrong, you will reject your own mail.

What to do next

  1. Lookup _dmarc on your domain with the DMARC checker. Confirm p= and rua= match what you think you published.
  2. Collect at least a business week of rua data before the first tighten.
  3. Fix alignment on your own tools before you raise p=.
  4. When the XML is the bottleneck, start a trial and point rua= at Email Watch (with authorization). The dashboard does not change p= for you; it tells you whether a change is safe.

Email Watch monitors. Receivers enforce whatever you publish in DNS.

Confirm the fix

Run the free check to confirm

Applied the fix above? Run the freeDMARC Checkeragainst your domain and see your grade update in real time.

Run the free check: DMARC Checker

Common follow-up questions

What does DMARC p=none do?

Receivers still authenticate and can send you aggregate reports, but they do not quarantine or reject based on your DMARC policy. Other filters (spam, reputation, bulk-sender rules) can still divert or bounce mail. none is a monitoring policy, not a promise that mail will land.

What is the difference between quarantine and reject?

quarantine asks the receiver to treat failing mail as suspicious, usually the junk folder. reject asks the receiver to refuse the message during SMTP. Both apply only to mail that fails DMARC (no aligned SPF or DKIM pass). Receivers may still soften policy.

How long should I stay at p=none?

Until you can name every high-volume source, legitimate senders show aligned SPF or DKIM, leftover failures are low volume or clearly spoofing, and you have watched a full sending week (including invoices, marketing, and tickets). A quiet Tuesday is not enough if a batch sender only runs on Monday.

Can I skip quarantine and go straight to reject?

You can publish reject at any time. The risk is a forgotten tool disappearing for everyone at once. Quarantine, or reject with pct= ramped up, turns that into a partial incident you can see in reports before it is total.

Does p=none satisfy Google and Yahoo bulk-sender rules?

Major providers have required working SPF and DKIM, and a published DMARC record, for bulk senders since 2024. That record can be p=none. Enforcement is separate from 'has a DMARC record at all.' Check current provider docs for volume thresholds; they change.

What are pct and sp?

pct= is the percentage of failing mail the policy should apply to (default 100). Use it to roll quarantine or reject in gradually. sp= is the policy for subdomains when they do not have their own DMARC record. Set it deliberately if you send from subdomains.

If I publish p=reject, will Email Watch block the mail?

No. Email Watch ingests reports. Receivers apply the policy in DNS. Changing p= in the _dmarc TXT record is what changes handling, with reporting lag of about a day.

How do I know it is safe to tighten?

Read rua reports (or a dashboard that parsed them) until unknown high-volume sources are gone and your own tools align. Confirm the live record with a DMARC checker before and after each change. Keep a copy of the previous record so you can roll back.