Project Server end of support: three paths

If you are running Project Server 2016 or 2019, the runway is already gone. Both versions reached end of extended support on July 14, 2026. Microsoft's lifecycle tables record an Extended End Date of 7/15/2026 6:59:59 AM Pacific for each — that is the timestamp boundary, so July 14 is the last full day. There is no Extended Security Updates program for Project Server at any price.

That much is a normal end-of-support story, and if it were the whole story the answer would be the usual one: pick the cloud destination and go. It is not the whole story, and the reason is the part most teams discover in week three.

The destination most teams name first retires before you finish

Ask a team what replaces on-premises Project Server and the first answer is almost always Project Online. It is the obvious name. It is the Microsoft-branded cloud version of the thing they already run.

Microsoft's lifecycle listing gives Project Online a retirement date of September 30, 2026. Project Server 2016 and 2019 died on July 14, 2026. The destination has a shorter remaining life than the thing you are leaving — by roughly two and a half months.

This is not a subtle trap. It is a published date on a public lifecycle page, and it inverts the entire migration plan. A Project Server–to–Project Online project scoped in good faith in the spring would have landed its users on a platform that retires before the change management finished. Anyone still holding that plan should stop and re-scope, because the second migration costs more than the first — you pay for the move twice and spend the credibility once.

Path one: Planner

Microsoft's forward path for work and project management is Planner, into which it has consolidated Project for the web, the older Planner, and To Do. For most mid-market organizations coming off Project Server this is the right destination, for the ordinary reason: it is where the vendor's investment is going, and it is the only option on this list with no published end date attached to it.

The honest caveats. It is not a like-for-like replacement for a mature Project Web App — the scheduling engine, enterprise custom fields, resource-pool semantics and timesheet model are not the same, and a portfolio built around Project Server's reporting contract will not lift cleanly. Licensing is its own conversation: what Planner Plan 1 buys versus Planner and Project Plan 3 decides whether the destination can actually hold your use case. Establish that before the migration, not after, because it is the question that most often reopens a closed scope.

Path two: Project Server Subscription Edition, on-premises

Staying on-premises is a real path, and for organizations with genuine data-sovereignty or regulatory constraints it is sometimes the only one. Project Server Subscription Edition is supported and is the on-premises continuation.

But price it honestly, because it is not a Project decision. Project Server ships inside the SharePoint Server Enterprise installer. It is not a separate product with a separate install; its files arrive in the SharePoint Enterprise MSI, and it can only be enabled on the Enterprise edition — never Standard. Project Server Subscription Edition requires SharePoint Server Subscription Edition Enterprise underneath it.

So this path is a SharePoint Enterprise licensing and upgrade project wearing a Project-shaped label. And SharePoint Server 2016 and 2019 went out of support on the same day, July 14, 2026 — which means the farm underneath your Project Server is in exactly the same position. If the database tier is SQL Server 2016, that went out of support that day too. Three tiers, one date. Scope all three together or the third one arrives as a change order.

Path three: off the Microsoft PPM stack

The third path is leaving. If your organization's Project Server usage is really scheduling and portfolio reporting rather than deep Microsoft integration, the migration budget you are about to spend is also a reasonable moment to ask whether a third-party portfolio platform fits better. That is a genuine option and we will say so when it is the right answer — we take no resale margin on any destination in this post, including Microsoft's.

The test is integration depth. If Project data feeds Power BI models, Entra-governed permissions and SharePoint-hosted project sites that the business actually uses, leaving is expensive. If Project Web App is effectively a scheduling tool with a resource pool attached, it is not.

Don't forget the desktop, which has its own date

The server is not the only surface with a 2026 date on it. Project 2021 and Project 2021 (LTSC) both end on October 13, 2026 — and they get there by different mechanisms, which is worth understanding because it changes what "end" means.

• Project 2021 follows the Modern Lifecycle Policy and carries a Retirement Date of 10/14/2026 6:59:59 AM Pacific.
• Project 2021 (LTSC) follows the Fixed Lifecycle Policy and reaches its Mainstream End Date on the same timestamp — with no Extended phase listed at all.

For the LTSC build in particular, there is no extended-support tail to fall back on. Mainstream end is the end. Teams that bought LTSC precisely to avoid a subscription are the ones most likely to assume an extended phase exists here, and on this product it does not.

What to do this month

The first task is not a migration, it is an inventory — and it is smaller than it sounds.

1. Establish which surface you are actually on. Project Server 2016, Project Server 2019, Project Online, Project Professional on the desktop, or Planner. Organizations frequently run two of these at once and have only budgeted for one.
2. Write down which date binds each one. July 14 2026, September 30 2026 and October 13 2026 are three different clocks on three different surfaces, and they are not interchangeable.
3. Check the tiers underneath. The SharePoint farm and the SQL instance. On this stack they usually share the July 14 date, and finding that out late is what turns a Project migration into three.
4. Inventory the customization tail before estimating. Custom fields, lookup tables, enterprise calendars, timesheet configuration, and the permission mode the farm is in. This is the scheduling variable, not the project count.
5. Then pick the destination — and check the destination's own lifecycle date before you commit to it. That single step is what this whole post exists to make routine.

Sources and further reading

• Microsoft Lifecycle — Project Server 2016 (Extended End Date 7/15/2026 6:59:59 AM PT)
• Microsoft Lifecycle — Project Server 2019 (Extended End Date 7/15/2026 6:59:59 AM PT)
• Microsoft Lifecycle — Project Online (Retirement Date 9/30/2026, Modern Lifecycle Policy)
• Microsoft Lifecycle — Project 2021 (Retirement Date 10/14/2026, Modern Lifecycle Policy)
• Microsoft Lifecycle — Project 2021 (LTSC) (Mainstream End Date 10/14/2026, Fixed Lifecycle Policy, no Extended phase)

The 30-second version

Project Server 2016 and 2019 ended support on July 14, 2026 — that date has passed, and there is no ESU program. Do not migrate to Project Online: it retires September 30, 2026, before the thing you are leaving finishes dying. Three real paths — Planner for most, Project Server Subscription Edition where sovereignty demands it (priced as the SharePoint Enterprise decision it actually is), or off the Microsoft PPM stack entirely. The desktop has its own date: October 13, 2026, and Project 2021 (LTSC) has no extended phase behind it. Inventory the surface, the date that binds it, and the tiers underneath — then pick a destination and check its lifecycle date first.

If you'd like a senior engineer to walk through it with you, the project intake form takes about three minutes. Two-business-day response with scope and a fixed-fee range.

Pro IT NW does senior-led Microsoft project work. Vendor-neutral. Labor-only. Based in Seattle, delivered USA-wide. We don't take resale margins on Microsoft licensing or on any of the destinations in this post.