Table of Contents
- What 550 5.7.708 Means in Exchange Online
- Four fields carry the diagnosis
- Telling a 550 5.7.708 Block Apart From Other Rejections
- What Triggers a Microsoft 365 Tenant Block
- What We Check First, in Diagnostic Order
- How to Open a Microsoft Support Case for 550 5.7.708
- What the reviewer is reading for
- Common Mistakes That Make 550 5.7.708 Worse
- Controls That Keep a Tenant Out of This Block
- Answers to Common Questions About 550 5.7.708
- Does 550 5.7.708 mean my domain is blacklisted?
- Is 550 5.7.708 the same as a temporary deferral?
- What if the block returns after Microsoft closes the case?
- Will switching to a new tenant clear the block?
- Decide Whether to Escalate or Bring In Help
Do not index
Do not index
550 5.7.708 means Exchange Online refused the message because it rates the sending IP as low reputation, and sending the same message again from the same IP will not clear it. The rejection belongs to the source, not to the recipient address, so the work starts with the tenant, its licensing, and its outbound path.
The pattern is familiar. Password resets stop arriving, invoices queue up, a support inbox fills with tickets, and the bounce folder shows the same line: "550 5.7.708 Service unavailable. Access denied, traffic not accepted from this IP." Marketing points at the ESP, IT points at the list, and the campaign keeps retrying into a wall that has nothing to do with either.
What follows is the order a deliverability team works the problem in. What the code means, how to separate it from every other bounce in the same folder, which tenant behaviors trigger it, the checks that come first, how to build a support case Microsoft can act on, and the controls that stop it from coming back.
Table of Contents
What 550 5.7.708 Means in Exchange OnlineFour fields carry the diagnosisTelling a 550 5.7.708 Block Apart From Other RejectionsWhat Triggers a Microsoft 365 Tenant BlockWhat We Check First, in Diagnostic OrderHow to Open a Microsoft Support Case for 550 5.7.708What the reviewer is reading forCommon Mistakes That Make 550 5.7.708 WorseControls That Keep a Tenant Out of This BlockAnswers to Common Questions About 550 5.7.708Does 550 5.7.708 mean my domain is blacklisted?Is 550 5.7.708 the same as a temporary deferral?What if the block returns after Microsoft closes the case?Will switching to a new tenant clear the block?Decide Whether to Escalate or Bring In Help
What 550 5.7.708 Means in Exchange Online
550 5.7.708 is Exchange Online's answer for a sender it does not trust. Microsoft documents the code as "Access denied, traffic not accepted from this IP," generated when the sending IP address has a low reputation, per the Microsoft troubleshooting guide for NDR codes 550 5.7.703 through 550 5.7.750. It is a policy rejection about the source, not a bad recipient address and not a temporary delay. Retrying from the same IP generally changes nothing, because the decision is tied to Microsoft's assessment of where the traffic came from.

Here is what it looks like when it lands in a mailbox, with the addresses replaced by example.com:
From: Mail Delivery System <postmaster@example.com>
To: campaigns@example.com
Subject: Undeliverable: Your receipt from Example Co
Reporting-MTA: dns; mx.example.com
Received: from mail.example.com [203.0.113.10]
by mx.example.com with ESMTPS
Diagnostic-Code: smtp; 550 5.7.708 Service unavailable.
Access denied, traffic not accepted from this IP
Remote-MTA: dns; outlook-com.olc.protection.outlook.com
Date: Mon, 6 Oct 2025 09:14:22 +0000 (UTC)Four fields carry the diagnosis
- The enhanced status code. 5.7.708 places this in the policy family, not the addressing family. A 5.1.x code is a different conversation.
- The status text. "Traffic not accepted from this IP" names the object under judgement. It is the sending IP, not the From domain.
- The reporting server. A remote MTA in Microsoft's outbound protection naming itself as the rejecting host confirms you are dealing with Exchange Online, not a recipient's own filter.
- The timestamp in UTC. This anchors the message to a minute in the tenant's sending history, which is how you find the volume spike or the first spam complaint.
The 5.7.7xx range is important context. Those codes come from Exchange Online and describe the sending tenant or source itself, which is why Microsoft's documentation for 5.7.700 to 5.7.750 is the reference guide here. A block issued against an IP outside Microsoft 365 arrives with different wording and a different escalation route.
Microsoft states the condition is more likely to affect new customers, including organizations using a Microsoft 365 trial subscription. A newly created tenant, a freshly provisioned sending path, or infrastructure with no established history can hit the block before it has built any trust. The party responsible for the fix is the email administrator, not the end user and not the Outlook client. Before going further, settle three facts: which system is rejecting, which IP Microsoft sees, and whether the tenant is new or on trial.
Telling a 550 5.7.708 Block Apart From Other Rejections
Start by confirming which block you have. Then match the status code and text to the row.
Bounce signal | Enforcement layer | Retry helps? | Next action |
550 5.7.708 Access denied, traffic not accepted from this IP | Exchange Online, judging the sending IP and tenant reputation | No, the same IP and message are rejected again | Run the tenant checks, then open a Microsoft support case |
4xx temporary deferral | Receiving server rate limit or greylisting | Yes, after the server's stated wait | Slow the send rate and retry later |
Block text naming an external sending IP | A blocklist for an IP outside Microsoft 365 | No, until the listing is removed | Delist through the list operator's own delist portal |
550 5.1.1 User unknown | Recipient address does not exist | No, and never | Remove or correct the address |
The trade-off is often misread. A 4xx deferral is the receiving server asking for time. It clears once the sending rate drops. A 550 5.7.708 is a trust decision already taken about the sending tenant and the IP Microsoft sees, so waiting does nothing. Retries of the same message from the same IP only repeat the pattern the filter already recorded. Lifecycle and outbound teams lose days when they treat that permanent block like a temporary deferral.
An external IP listing is a separate matter. Microsoft documents that the reputation object can be a relay, a shared gateway, a marketing platform, or a newly changed connector, so the mailbox the sender sees may not be what Exchange Online is judging, as the Microsoft Learn guidance on the 5.7.700 to 5.7.750 codes sets out. A listing held by the list operator comes off at that operator's delist portal, not through a Microsoft ticket.
Read the whole bounce, not only the first line. Microsoft sometimes prints an AS code or an outbound reference next to the status. A 550 5.1.1 still means the recipient address does not exist, and sending it again never helps. If the text is ambiguous, treat it as a 5.7.708 and let the tenant checks rule it in or out.
What Triggers a Microsoft 365 Tenant Block
A tenant can start receiving 550 5.7.708 after a single change, such as launching a new campaign or routing mail through a different connector. The code points to a tenant-level trust decision, so investigate who sent the mail, how it was sent, and which system carried it.
- A new or trial tenant sending real volume. A tenant without established sending history may be blocked when its first campaign sends at scale. Cost: the launch stops before engagement data can guide adjustments.
- A sudden jump from a quiet tenant. A dormant mailbox sending a product announcement can resemble compromised activity. Cost: the announcement and transactional mail using the same route may stop.
- A compromised account sending spam or phishing. An account takeover can expose the tenant to reputation damage and suspension. Reviewing account security and controls that detect and prevent mailbox hijacking helps reduce this risk. Cost: recovery can require a support case and reputation repair.
- Bulk or cold outreach from a new tenant. Sending to purchased or scraped lists can generate bounces and complaints. Cost: the block can halt outreach while damaging the sending domain's reputation.
- Weak list hygiene. High bounce rates and unknown addresses suggest that recipient data needs attention. Cost: filtering can restrict delivery before recipients see the message. A reliable email list building process helps keep invalid and unknown addresses off the file.
- An open or shared relay. A misconfigured connector can let third parties send through the tenant. Cost: the tenant may absorb reputation harm from mail its team did not intend to send.
- A changed sending path. A new connector or platform-side relay can change the IP Microsoft sees. Cost: investigating the wrong sending address delays identifying the source of the problem.

A block can affect more than campaigns. If password resets and invoices share the blocked tenant or sending route, customers may be unable to access accounts or receive billing messages, creating support work and lost revenue.
What We Check First, in Diagnostic Order
Microsoft's guidance for trials says an administrator may need to contact Microsoft Support for an IP exception after Exchange Online licenses are assigned, and the Exchange documentation adds a harsher reading: code 708 can indicate most traffic from the tenant is considered suspicious, potentially ending the tenant's ability to send. Both are documented in Microsoft's 5.7.703 to 5.7.750 troubleshooting article. That is why the checks run in this order. Each one either rules out a bigger problem or hands the next check its evidence.
- Preserve the evidence. Capture the full NDR, the SMTP transcript, the UTC timestamp, the recipient, the sender, and the apparent outbound IP. Without this, every later step is guesswork.
- Identify the enforcement layer. Decide whether the rejected IP sits inside Microsoft 365 or belongs to a gateway, CRM, marketing platform, or self-hosted relay. A mistake here sends the ticket to the wrong company.
- Confirm tenant state. Check whether the tenant is new or on trial, whether Exchange Online licenses are assigned to every sending account, and whether the sending domain is verified. Unlicensed or unverified senders produce exactly this class of failure.
- Hunt for compromise. Look for unfamiliar sign-in locations, new forwarding rules, and sent items the user didn't write. Reset passwords, revoke sessions, and enforce MFA. Compromise comes before policy tuning because a live intruder defeats every other fix.
- Review the outbound spam policy and volume history. Compare recent sending against the tenant's normal baseline. A step change is the trigger; a flat line points elsewhere.
- Run an Exchange message trace. Trace the rejected messages to confirm the actual path and the egress IP rather than the one in the diagram.
- Check the IP against blocklists. Use the free blacklist checker from our sister brand mailX and, if the address belongs to another provider, escalate with that provider.

Ownership is where most of these incidents stall. Security investigates compromise and open relays, IT verifies licensing and tenant state, and the sending platform confirms the egress IP. Authentication belongs in the same pass: check SPF, DKIM, and DMARC alignment in the headers so the case can show the block isn't an authentication failure, even though fixing records alone won't lift an IP-level restriction. The MailAdept guide to email authentication covers what to verify and why.
How to Open a Microsoft Support Case for 550 5.7.708
A case works when it reads as a closed incident. Before opening one, secure any compromised account, close any open relay, pause bulk sending, and confirm that the tenant is licensed and the sending domain is verified. The case should explain what caused the restriction and what has changed.
- Assemble the evidence. Include two or three representative NDRs in full, the SMTP transcript, UTC timestamps, recipient and sender addresses, the apparent outbound IP, and the message trace output.
- Add a day-by-day volume history for the preceding weeks. Mark when the sending pattern changed so the reviewer can compare the traffic history with the incident.
- Explain the corrective work. State what was compromised or misconfigured, what changed, which accounts and connectors were reviewed, and what sending will look like next.
- Open a request through the tenant's admin support channel and ask for a review of the IP or tenant restriction. Microsoft's trial guidance directs administrators to support for an IP exception once licensing is in place.
- Keep replies in one case thread. A single thread preserves the incident history and lets the reviewer assess new evidence in context.
What the reviewer is reading for
The reviewer needs evidence that the cause has been addressed. Explain the corrective work plainly, and attach evidence that supports the account of what happened. A secured account, for example, is stronger evidence than an unsupported claim that the traffic is legitimate. Add new information to the same thread.
For a trial tenant, confirm licensing before requesting a review, since the documented support path assumes licenses are assigned. Use the time before escalation to complete containment and organize the evidence; avoid promising a resolution date.
Common Mistakes That Make 550 5.7.708 Worse
Launching a new tenant or mailbox to keep the same campaigns running often repeats the problem. The new source has no established sending history, and the same traffic pattern can draw another block. It also creates a second tenant to investigate and clean up.
- Asking one recipient organization to allowlist you. A local allowlist cannot remove a Microsoft-wide restriction. It may help delivery to that organization while leaving the wider block untouched.
- Changing only SPF, DKIM, or DMARC and waiting. Authentication records matter, but passing those checks does not resolve a restriction based on the tenant's sending behavior.
- Switching sending domains while a compromise is active. Changing the domain does not contain the source of the traffic. Secure the affected accounts and sending systems before using another domain.
- Warming an IP at volume during an active block. Increasing traffic while Microsoft is rejecting it adds more failed sends without establishing that the underlying issue is resolved.
Use a clear sequence: stop affected sends, secure the tenant, document what changed, and escalate through the support case. Avoid workarounds that create more sending activity before the cause has been contained.
Controls That Keep a Tenant Out of This Block
Control | Why it matters | What happens if ignored |
Ramp volume gradually on new or trial tenants | Builds sending history before peak volume | A sudden spike from a new tenant can trigger a low-reputation rejection |
Separate transactional and promotional streams | Protects password resets and receipts from campaign damage | One bulk campaign can block the mail customers need most |
Enforce MFA and review sign-ins for every sending account | Helps prevent compromised accounts from sending spam | A compromise can lead to suspension and a support case |
Keep hard bounces low. Under 2% is a practitioner rule of thumb, not an official Microsoft limit. | Reduces the risk of sending to unknown and invalid addresses | Bounce spikes can damage reputation |
Keep complaints low. Gmail and Yahoo's bulk sender requirements include a complaint-rate threshold; see their official bulk sender guidelines. | Helps keep complaint rates within the published requirements | Complaint spikes can put every stream at risk |
Review outbound volume and bounces weekly, with a named owner | Catches patterns early | Incidents can recur without anyone accountable |
These controls work together. A promotion sent to a stale list can produce both bounces and complaints, increasing reputation risk across the sending stream. Cold outreach has similar list-quality and volume concerns, which the MailAdept guide to cold email deliverability addresses.
Protect the sending tenant with clear ownership across security, IT, and the team responsible for the sending platform. Review access and sending activity regularly, and keep customer-facing transactional mail separate from promotional campaigns. Domain safeguards matter too: block lookalike domain spoofing to reduce the risk of forged sender identities triggering complaints.
Review the sending setup after a new platform connection, connector change, or relay addition. These changes can alter the IP Microsoft sees, even when the team expects mail to use a familiar route. For high-volume sending to Outlook.com, include Microsoft's published sender requirements in the review.
Answers to Common Questions About 550 5.7.708
Does 550 5.7.708 mean my domain is blacklisted?
No. The rejection can reflect Microsoft's trust assessment of the sending IP or tenant, even when the domain itself is clean. Check the IP and any relay used by your sending platform alongside the domain.
Is 550 5.7.708 the same as a temporary deferral?
No. A 5.7.708 response is a policy rejection, not a routine delay. See the earlier comparison with 4xx deferrals for how the two differ.
What if the block returns after Microsoft closes the case?
Treat the recurrence as a new signal to investigate, not as a reason to retry the same traffic. Confirm that the original cause remains resolved, then check whether a connected account, sending platform, relay, or recent configuration change has introduced a new risk. Assign an owner who can coordinate security, IT, and the sending platform, and bring Microsoft the new rejection details and a summary of what changed.
Will switching to a new tenant clear the block?
Do not assume it will. A new tenant does not resolve the underlying trust issue if the same risky sending practices or infrastructure carry over. Identify and address the cause before moving traffic.
Decide Whether to Escalate or Bring In Help
If the same rejection continues after the account is secured, the tenant is licensed, the domain is verified, and bulk sending has stopped, open a Microsoft support case. Continued retries add no useful evidence.
Bring in help when:
- The same code returns after a support case has closed.
- A migration to a new tenant, domain, or sending platform is planned or underway.
- Blocks affect multiple streams, such as transactional and marketing mail.
- No single owner spans security, IT, and marketing, leaving nobody able to coordinate the investigation.
MailAdept's Fix step identifies and resolves technical issues, reputation risks, and infrastructure misconfigurations. Its subscription service pairs a dedicated deliverability expert with the client's team, with optional implementation and monitoring support. Plans include calls, 12 or more daily checks per domain, and weekly expert reviews. For an urgent incident, MailAdept's deliverability consulting can support the investigation. If the issue also involves wider IT and mail infrastructure, the We Fix PC Laptop partner guide can help frame the managed IT support conversation.
Name an owner across security, IT, and the sending platform before the next step. That person can coordinate evidence, remediation, and the Microsoft case. For support with the deliverability investigation, start with a free audit from MailAdept.
