News
The 2025-26 enrollment numbers steering the CIP
MCPS enrolled 156,050 students this year, down for the third straight year. Here's what's driving it and why the capital budget hinges on the number.
Last September, MCPS published its official enrollments for the 2025-26 year: 156,050 students. This is down about 3,100 (−2.0%) from the year before. That’s the third consecutive annual decline and the largest one-year drop since the COVID year, when enrollment fell roughly 4,300. The district peaked near 164,900 in 2019-20 and, even after these losses, is still the largest school system in Maryland.
A number like this rarely means one thing. We pulled it apart on our new enrollment deep-dive.
Two drivers are pulling the number down
Smaller kindergarten classes. Kindergarten is down 14% from its 2019-20 peak, tracking a decline in county births. This one is essentially locked in: the children who will start kindergarten through fall 2027 are already born. You can watch this cohort work its way up the grades for years.
Older students leaving. The county loses students within a grade cohort from one fall to the next — the losses concentrated in grades 9-12, about −900 in fall 2025, and heaviest at high schools with large immigrant and multilingual enrollment. Elementary schools still gain from families moving in, but only about a third as many as in the 2023 rebound.
For what it’s worth, MCPS’s own projections had anticipated a decline this year. They projected slightly low; independent modeling (including ours) projected slightly high. Reality landed between the two.
Where it goes from here
Before accounting for new housing, the models point to roughly 7,000 fewer students by 2030-31. Approved residential development is expected to add back around 2,700 on a probability-weighted basis — counting fully-approved projects near-term and not-yet-entitled phases much more cautiously — offsetting a bit more than a third: a net decline near 4,300 (−2.8%) over five years. And it doesn’t stop at 2031: the smaller birth cohorts keep aging upward.
These are scenarios, not predictions — the kindergartners of 2031 and beyond aren’t born yet, and approved development is not the same as poured concrete. We lay out the assumptions on the enrollment page.
Why this is the CIP’s problem now
This is the part that matters for the county’s spending. A falling countywide total is easy to misread as “we have empty seats, so stop building.” The board’s adopted March 2026 changes cut against that reading — and for good reason.
Reopening Charles W. Woodward High as a relief school in 2027-28, moving Wootton’s program into the new Crown Farm building, and adding on to Damascus Elementary were all justified by overcrowding that is real today and persists in pockets even as the county total falls. Several high schools are still projected severely over capacity into 2031 — Churchill, Blake, Richard Montgomery, Einstein among them — and some elementary areas, like Damascus and Clarksburg, stay tight because new development concentrates in exactly the places already crowded. Our school-level forecasts show which buildings those are.
The flip side is just as real. As the total drifts down, more schools will edge under-enrolled, and that raises a different set of questions: which renovations get priority, whether to right-size buildings, where to place programs, and whether planned capacity projects still pencil out. Growth and decline are happening in the same county at the same time, a few miles apart.
The timing is the point. The county’s capital budget runs on CIP amendments through the fall and winter, and the enrollment count published now is the number those decisions will lean on. A single countywide figure won’t settle any of it — but it frames every argument that follows.
What to watch
Two things. First, the fall CIP process — whether the adopted projects hold and how the board weighs crowded schools against emptying ones. Second, next year’s September 30 count. Our enrollment page states plainly what it expects for 2026-27, so when that number lands you can check the model against reality — the same test we just ran on this year’s.