Microsoft Exchange Server is still a supported on-premises messaging product, but the answer depends entirely on the release. In October 2026, an organization running Exchange Server Subscription Edition (SE) is in a different position from an organization still operating Exchange Server 2013, 2016, or 2019. Treating those releases as one undifferentiated product can lead to an unsafe upgrade plan.
The practical question is not simply whether Exchange still exists. It is which version is installed, whether that version receives updates through its applicable support arrangements, and what path leads to a supported platform without disrupting mail flow, authentication, retention, or business-critical applications. The older Exchange Server 2013 deployments that prompted many end-of-support questions remain especially important as historical migration cases; that release is not a current supported option.
Which Exchange Server versions are supported in October 2026?
Exchange Server 2013 reached its extended support end on April 12, 2023. Exchange Server 2016 and Exchange Server 2019 both reached their regular support end on October 14, 2025. Microsoft identifies Exchange Server Subscription Edition, introduced in 2025, as the current on-premises Exchange release under the Modern Lifecycle Policy. The fact that a server is still running does not establish support, security coverage, or compatibility with a future operating-system update.
Some customers enrolled in Microsoft’s time-limited Extended Security Update arrangements for Exchange Server 2016 or 2019 can receive specified security updates after ordinary support ended. That exception is narrower than regular product support: it does not make the older releases newly current, does not automatically restore feature development, and must be verified against the organization’s actual enrollment and Microsoft servicing eligibility. For an ordinary un-enrolled installation, plan replacement rather than assuming the historical release will continue to receive routine fixes.
Microsoft’s Exchange Server 2016 and 2019 lifecycle notice and the Exchange Server 2013 support record provide dated issuer context. Recheck these official notices before committing to a migration schedule, because servicing exceptions and prerequisite builds are time-sensitive.
Why continuing to operate an unsupported server is a business risk
An unsupported Exchange environment can retain every familiar mailbox and Outlook connection while the underlying attack surface grows. Remote-access endpoints, transport agents, administrative interfaces, Windows dependencies, certificates, and third-party backup components all change around it. A vulnerability that is patched for a supported build may remain exposed on an older build. Network isolation helps reduce exposure but does not turn an unsupported product into a supported one.
Security risk also extends beyond the server itself. Mail transport depends on DNS records, public certificates, anti-spam and anti-malware controls, identity systems, outbound connectors, and external business applications. An application using basic SMTP credentials or a hard-coded relay may fail when the messaging platform is migrated. Identifying these dependencies before making changes is more effective than diagnosing broken invoice delivery or scanner alerts after the move.
Separate three objectives: preserving delivery, protecting mailbox data, and demonstrating recovery. A healthy database and a successful backup are not proof that an organization can restore mail access within its target outage window. An existing business continuity and disaster recovery plan should establish service priorities, restoration responsibilities, dependency order, and the tests that demonstrate readiness.
Choose the destination before designing the upgrade
An organization may move to Exchange Online, deploy Exchange Server SE on premises, or adopt an appropriate hybrid design. The decision is driven by regulatory requirements, data residency, integration constraints, connectivity, operational capacity, and the cost of continuing to run messaging infrastructure. No destination automatically solves every requirement. A team with specialized mail-processing workflows must test those integrations; a team seeking less infrastructure maintenance may value a managed cloud service while accepting a different identity, compliance, and operating model.
Exchange Online migration is not simply a mailbox-export exercise. Verify supported identity and directory synchronization arrangements, the target tenants and domains, transport connectors, coexistence requirements, archive and retention settings, shared mailboxes, distribution groups, and how clients and mobile devices will authenticate. A migration that moves messages but breaks journaling, line-of-business relays, or delegated mailbox access is not complete.
For organizations that must remain on premises, Exchange Server SE changes the lifecycle discussion. It has its own installation prerequisites, servicing requirements, licensing considerations, and compatibility matrix. Confirm the exact operating-system, Active Directory, hardware and Exchange build combinations against Microsoft’s current supportability documentation. An otherwise sensible migration can stall at setup if the directory schema, coexistence restrictions, or prerequisite build is wrong.
Upgrade paths differ for Exchange 2013, 2016, and 2019
Microsoft documents an in-place Exchange Server SE upgrade from Exchange Server 2019 CU14 or CU15, subject to the applicable prerequisites and servicing path. That does not mean an installation can jump in place from every older Exchange version. Exchange Server 2016 generally requires a legacy migration that introduces the target server, transfers mailboxes and resources, and removes the old server; an in-place conversion straight from 2016 is not a supported shortcut.
Exchange Server 2013 needs particular care. Exchange SE’s original release does not support coexistence with Exchange Server 2013. Microsoft’s documented path involves bringing the organization to an appropriate later version first, fully removing Exchange 2013, and then completing the supported Exchange SE transition. Organizations that have not upgraded since 2013 should not improvise a direct installation into the production organization. Run the documented compatibility and coexistence checks before any schema or server change.
Microsoft’s Exchange Server SE upgrade procedures distinguish a legacy migration from an in-place upgrade and describe support limitations for older-server coexistence. This official page is the operational source for selecting the correct method; project timelines should be built around its current prerequisites rather than generic assurances of a seamless upgrade.
Moving to another server also creates a technical migration plan beyond Exchange setup. Inventory certificates, namespace records, virtual directories, receive and send connectors, mailbox databases, arbitration and system mailboxes, transport rules, supported Outlook clients, backup jobs, and monitoring hooks. During coexistence, changes to accepted domains or DNS should have an explicit rollback plan and a validation window. Plan account permissions and application secrets as carefully as mailbox content.
Test authentication, SMTP transport, and compliance before cutover
Directory and authentication health is a prerequisite for reliable messaging. Test name resolution, domain controller availability, certificates, access policies, and the precise authentication methods used by staff and applications. A technically migrated mailbox is not available to its owner if the identity path is broken. Enterprises with complicated legacy domain structures benefit from a separate Active Directory Domain Services review covering directory roles, replication, and security boundaries.
SMTP deserves its own workstream. Mailbox users are not the only senders: printers, monitoring products, workflow platforms, scanners, websites, and integration services may all depend on Exchange as a relay. Preserve or deliberately replace those flows. Port 25 ordinarily handles server-to-server SMTP transport, while authenticated client message submission commonly uses port 587 under a distinct security policy. The SMTP port 25 and 587 distinctions matter when testing connectors and setting firewall rules, but actual relay design depends on the deployed Exchange and network configuration.
Compliance controls require explicit inspection. Validate legal hold, archive retention, discovery/export procedures, mail routing through third-party security gateways, address-book visibility, and delegated access. If the organization must preserve historical messages for a fixed period, document not only the migration result but also chain-of-custody evidence, backup restoration tests, and who can access retained data.
Before cutover, run test messages through real business workflows, not just a send-to-self check. Validate inbound mail from multiple external providers, outbound relay through required connectors, attachments, calendar invites, meeting-room resources, mobile devices, shared mailboxes, automated alerting, and any message encryption or inspection systems. Capture expected outcomes and compare them after each migration phase.
Execute a staged migration with measurable rollback conditions
A manageable first phase is discovery. Export an accurate server and mailbox inventory, record build and patch levels, identify unsupported components, and establish baselines for transport queues, authentication failures, database health, mail volume, backups, and external delivery. Discovery often changes the plan: a small number of applications may account for most the migration complexity, even when the mailbox count is straightforward.
The second phase is a representative pilot. Select users and applications that exercise different features rather than only volunteers with simple mailboxes. Confirm end-user access, transport behavior, mobile enrollment, mail delegation, archives, and recovery. Keep the pilot small enough that the support team can diagnose failures, but representative enough to expose shared dependencies. Record unresolved problems and stop criteria before expanding the migration.
As groups are transferred, monitor queues, certificate warnings, Autodiscover behavior, Outlook connection state, identity failures, and external delivery. Keep change windows aligned with business use, particularly payroll, customer notifications, or support operations. A rollback plan needs exact trigger conditions, a known-good infrastructure baseline, ownership, and tests that can prove recovery. Returning DNS alone may not restore operations if mailboxes or routes have already changed.
The last phase is deliberate retirement. Decommission old servers using vendor-supported removal procedures and only after confirming mailbox, connector, transport and compliance dependencies are gone. Remove obsolete external exposure, revoke secrets, retire certificates as appropriate, update asset inventory and operational diagrams, and record support ownership for the new system. An unsupported Exchange server left powered on for an unidentified application becomes a future incident risk.
What administrators and certification learners should take away
The most important lesson is to distinguish product lifecycle from technical availability. A server can keep delivering email after support ends, but successful delivery does not demonstrate security support, compliance, or recoverability. Similarly, an Exchange SE deployment does not become resilient merely because it is newer; its directory, transport, access, patching, monitoring, and restore dependencies still require sound design.
When evaluating any Exchange environment in 2026, establish the exact version and build, confirm supported servicing, choose Exchange Online or Exchange SE on actual business requirements, and execute a tested migration that includes transport, identity, data, and rollback. Historical Exchange 2013 content remains useful for understanding the transition, but it must never be mistaken for a current recommendation to keep running an unsupported server.