Emails Missing After MX Record Update — what to check first (without breaking mail flow)

Updated: January 13, 2026 · Answered by Steve H.

Steve H.

Expert ★★★★★ 27,853 satisfied customers

Specialities: MX record change triage, DNS propagation/TTL validation, mail routing and priority conflicts, split-delivery detection, safe inbound mail diagnostics.

After an MX record update, missing email is usually DNS propagation, wrong authoritative zone, MX priority conflicts, or old routes still receiving mail — here’s what to check first

Technician

Hi! “Emails missing after MX record update” is usually not data loss — it’s almost always mail being delivered somewhere else (old provider, security gateway, or a cached DNS path).

Let’s keep this safe: I won’t ask you to delete records, “wipe” anything, or keep flipping settings. We’ll start with the least risky checks and build evidence.

Quick basics so I can triage correctly:
1) What email platform are you moving to (Google Workspace, Microsoft 365, cPanel, Exchange, mail gateway)?
2) What exactly is “missing” — inbound from everyone, only from some senders (Gmail/Outlook), or only to certain recipients?
3) When did you change the MX records, and do you know the previous TTL?

Customer

We changed MX yesterday evening to Microsoft 365. Some emails arrive, but others never show up. People say they sent them. No bounce messages. I’m worried we broke inbound mail.

Technician

Understood — the “some arrive, some don’t” pattern is classic for DNS propagation + caching or split delivery (different senders hit different MX answers). That’s why you may not see bounces: mail can still be accepted — just not by the place you’re checking.

Two safe questions (no changes yet):
1) Where was mail hosted before (old mailbox provider or an email security gateway like Proofpoint/Mimecast)?
2) Do you still have access to the old system to confirm whether it’s receiving any new inbound mail?

Customer

Previously it was a cPanel mailbox on our web host. Yes, we still can log into those old mailboxes. I haven’t checked them today. We just assumed everything moved.

Technician

Great — that gives us a safe “control group.” First trust moment: I don’t want you to delete the old mailboxes yet. During propagation, they often keep receiving mail for a while.

Safe check #1 (takes 2 minutes): open one of the old cPanel inboxes and see if any of the “missing” messages landed there. If yes, we’re dealing with split routing, not missing mail.

Next, we’ll confirm what the internet sees for your domain’s MX. Can you tell me the domain name (just the domain, not any passwords), and whether your DNS is managed at the registrar, Cloudflare, your host, or somewhere else?

Customer

DNS is at Cloudflare. Domain is example-company.com. And wow… I checked the old cPanel inbox and there are new messages from this morning that aren’t in Microsoft 365.

Technician

That’s actually good news — it means mail isn’t “vanishing.” It’s being delivered to two different destinations depending on which resolver/sender is asking DNS.

Common reasons after an MX update (senior-friendly but precise):
• Propagation/TTL: some senders still cache old MX answers.
• Wrong zone edited: MX changed in Cloudflare, but the domain may still be using different authoritative nameservers (or a second DNS provider is active).
• MX conflict: old MX records still exist alongside the new one, or priorities are wrong.

Safe check #2: In Cloudflare DNS, do you see only the Microsoft 365 MX record(s), or do you also still see the old cPanel/host MX entries?

Customer

I see one MX for Microsoft (priority 0) and also an older MX that points to mail.example-company.com with priority 10. I left it “just in case.” Is that bad?

Technician

It’s understandable to keep a fallback — but yes, that setup can cause exactly what you’re seeing. Many senders will try the lowest number first, then may fall back to the next MX if there’s any temporary delivery problem. That can create “random” delivery between Microsoft 365 and the old host.

Second trust moment: I’m not telling you to delete it yet in the middle of an incident. First, we should confirm the current public answer and whether the domain is truly using Cloudflare as the authoritative DNS.

Safe check #3 (no changes): can you run one lookup from your side and paste only the results (no private info):
• nslookup -type=mx example-company.com (Windows)
or dig mx example-company.com (Mac/Linux).

If you can’t run commands, tell me what the MX record hostname is for Microsoft 365 that you entered (it usually looks like example-company-com.mail.protection.outlook.com).

Customer

I can run it. It returns two MX records: the Microsoft one and the mail.example-company.com one. Also, some colleagues on different networks see different results in online MX checkers.

Technician

That confirms split routing. The “different results” across checkers can happen when they query different resolvers and caches, especially within the first 24–48 hours after a change (TTL matters).

Now we want to answer one key question: Is Microsoft 365 receiving the mail but not delivering it, or is the mail never reaching Microsoft 365 at all?

Safe check #4 (no risky edits): In Microsoft 365, do you have access to the Exchange Admin Center message trace (or any inbound logs)? Even a simple “did the message hit Microsoft at all” test helps. If you don’t have admin access, tell me whether you’re using a connector, a third-party spam filter, or any inbound gateway.

Customer

We have message trace. For the emails that “disappear,” there’s nothing in the trace. For the ones that arrive, trace is normal. We do not use a third-party gateway.

Technician

Perfect diagnostic — if it’s not in message trace, those messages likely never hit Microsoft 365. They’re going to the other MX (the old host) or being retried elsewhere based on cached answers.

Third trust moment: I don’t want you to “fix” this by repeatedly changing records back and forth. That usually makes propagation chaos worse. Instead, we do controlled checks and one clean cutover plan.

A few quick checks that senior admins usually overlook during MX cutovers:
• Is the old MX host (mail.example-company.com) still valid and accepting mail? (It is, based on your cPanel inbox.)
• Does your domain have multiple DNS providers by accident (registrar + Cloudflare)? Check that the domain’s nameservers at the registrar are Cloudflare, and only Cloudflare.
• Did you change TTL shortly before the update, or was it high (hours/days)?

Customer

TTL was 1 hour for MX. Nameservers are Cloudflare, yes. We didn’t change anything at the registrar besides that. So the big issue is that extra old MX still exists, right?

Technician

Most likely, yes: keeping two MX endpoints during a cutover commonly creates “some messages missing” because senders can legitimately deliver to either path depending on their retry behavior and cache.

But before any change, one more safe validation (this prevents a worse outage): confirm the Microsoft 365 MX value is exactly correct and that there are no typos or wrong target domain. Also confirm there isn’t an MX record for a subdomain that users are accidentally sending to (like mail.example-company.com vs example-company.com).

Can you confirm: are users emailing @example-company.com only, not a subdomain? And did you also update SPF/DKIM/DMARC recently, or only MX?

Customer

Only @example-company.com. We also updated SPF to include Microsoft, but DKIM/DMARC haven’t been touched yet. We’re not seeing rejects though.

Technician

Good — SPF/DKIM/DMARC changes usually affect deliverability and spam placement, but they don’t typically make mail “disappear” without any trace. Your symptom (no Microsoft trace, mail shows up in old inbox) points back to MX routing.

Safe action plan (still not telling you to click “delete” yet):
1) Identify 2–3 missing messages and confirm where they landed (old cPanel vs Microsoft).
2) Check MX answers from at least two resolvers (your local + a public one) and record what they return.
3) Confirm there’s only one intended inbound path for the primary domain during the cutover window.

Once you have that, you can do a controlled change and then monitor message trace + the old inbox for stragglers.

Customer

That makes sense. My fear is making it worse. If I remove the old MX and something is wrong with Microsoft MX, we could lose everything.

Technician

That’s a smart concern — and it’s why we validate first.

The safe approach is:
• Verify correctness of the Microsoft 365 MX target (exact hostname) and that you can see successful inbound delivery in message trace for test mail.
• Keep a short observation window after any change where you monitor both Microsoft message trace and the old cPanel inbox for any remaining deliveries.
• Avoid rapid changes — one controlled update, then wait for caches to expire based on TTL.

If you follow that, you won’t “lose everything.” At worst, some mail will keep hitting the old inbox until caches expire — which you can watch and then consolidate once routing is stable.

Customer

Got it. I’ll document where the missing emails are landing, confirm the exact MX value, then do one controlled update and monitor. This is way calmer than random guessing.

Quick answer: the accepted fix

Accepted approach when emails go missing after an MX record update: confirm whether messages are landing in the old inbox (split routing), verify you edited the authoritative DNS zone (correct nameservers), and check that only the intended MX records exist with the correct priority. Then validate inbound delivery using message trace/logs (so you know whether mail reaches the new provider at all). Make one controlled DNS change and monitor both destinations until caches expire (TTL), rather than flipping settings repeatedly.

Why email can “disappear” right after changing MX records

After an MX update, different senders may still use cached DNS answers for hours (or longer if the prior TTL was high). That creates split delivery: some mail reaches the new provider, while other messages continue going to the old mail server. Another common cause is editing the wrong DNS place (not the authoritative nameservers), or leaving multiple MX records in place so senders can legitimately deliver to more than one route.

Safe checks before you change more DNS settings

Start with reversible diagnostics: check the old mailbox for new inbound messages, confirm the domain’s nameservers match your DNS host, and verify the public MX answer using nslookup/dig from more than one network. Use message trace (Microsoft 365 / Google Workspace logs) to confirm whether missing emails ever reach the new provider. Avoid multiple rapid MX changes — controlled updates + careful monitoring usually resolve “missing mail” without creating a larger outage.

Safety note: don’t shut down the old inbox during propagation

This is how teams accidentally “lose” mail during a cutover.

Keep the old mailboxes accessible until you can confirm (1) the new MX is correct, (2) inbound mail is visible in the new provider’s logs, and (3) the old system stops receiving new messages after caches expire. If you remove the old system too early, messages delivered to the old route during propagation can be hard to track down later.

Disclaimer: This information is AI-generated and intended for general guidance only. Steps vary by provider and DNS host. For exact instructions for your setup, consult a verified expert or the manufacturer’s support site.

Understand what these prompts mean

Getting Started Is Easy

Low-risk diagnosis

Confirm whether mail is hitting the old route, the new route, or both — using safe checks like MX lookups and message trace.

Reduce cutover risk

Understand how TTL and cached resolvers cause split delivery, so you don’t “fix” it by making DNS worse.

Escalation-ready notes

Capture resolver results and provider logs so a senior admin (or vendor support) can act quickly and safely.

FAQ

Why did it start suddenly right after the MX record update?

Because not every sender updates at the same time. Some systems keep using cached DNS answers based on TTL, so mail can still route to the old server for a while. If you also left the old MX record in place, some senders may legitimately deliver there as a fallback. Checking both destinations is the fastest way to confirm this.

Emails missing after update — how long can DNS propagation take?

It depends on the previous TTL and how long external resolvers cache results. Even if you change records instantly at your DNS host, some senders may use old MX answers for hours. During this window, you can see split delivery. Monitor message trace/logs and the old inbox until routing stabilizes.

Is this a hardware problem or just DNS/settings?

Almost always DNS/settings. Hardware failures typically cause clear bounces or consistent failures, not “some emails arrive, some don’t.” When messages never appear in the new provider’s logs, it’s usually routing (MX, priorities, authoritative zone). Checking public MX answers and provider logs is the safest way to confirm.

Can this be fixed without calling a service or vendor?

Often yes, if you approach it methodically. Validate the authoritative nameservers, confirm the correct MX target and priority, and verify inbound delivery using provider logs. Avoid multiple rapid DNS changes and keep the old inbox available during propagation. If results are inconsistent after caches should have expired, then escalation may be appropriate.

Sources

Filed under: Account Access Help

Steve H.

After an MX record update, missing email is usually DNS propagation, wrong authoritative zone, MX priority conflicts, or old routes still receiving mail — here’s what to check first

Quick answer: the accepted fix

Why email can “disappear” right after changing MX records

Safe checks before you change more DNS settings

Safety note: don’t shut down the old inbox during propagation

Getting Started Is Easy

Low-risk diagnosis

Reduce cutover risk

Escalation-ready notes

FAQ

Sources

Related articles