Medical office after 2025. It is half as empty as office, it costs a third less to rent, and almost nothing new is being finished.
A lender or a fund that treats medical office as a kind of office will misread it. We read the third-quarter 2026 brokerage reports for both, set them side by side, added five metros, the cap rate readings and the supply figures, and looked for what is known about loans coming due. The last part is thin, and the page says where.

In the third quarter of medical outpatient buildings had vacancy against for office, a gap of . Rents are lower, against per square foot, and the pipeline is thin: just was finished in the quarter, against just over a million a year earlier. Investors have noticed. Cap rates run from depending on who is buying, and trailing-year sales are up . What we could not find is a loan maturity schedule for the type.
Medical outpatient space is 8.5 points less empty than office, and rents have kept rising.
The brokerage’s third-quarter 2026 report for medical outpatient buildings puts vacancy at , down on the year, with of net absorption. Asking rent is , up . The same brokerage’s report for all office has vacancy at , and asking rent at .
The gap is of vacancy. Medical rent is office rent. Cheaper space that fills faster is the first reason the two trade differently. A second source reports renewal rates above , so tenants who are in place mostly stay.
Vacancy fell 60 basis points in a year to 18.2 %, and prime space is at 11.9 %. Asking rent rose 3.1 % to $37.95. Net absorption was 4.9 million sq. ft., a tenth straight positive quarter.
Both columns are from the same brokerage for the third quarter of 2026. The office figure covers all U.S. office, the medical figure covers medical outpatient buildings.
The second-source figures are not the same series. One broker puts national occupancy in the top metros at , flat on the quarter, which is a vacancy of on a narrower base. The two readings agree on direction. Neither is a measure of your building.
A national average hides the metro spread. Five metros show how wide it is.
Vacancy runs from 5.3 % to 11.8 % across five metros, and Austin charges the most for it.
Houston is the loosest of the five at vacancy (derived from occupancy), with and close behind. Columbus is at (derived from occupancy) and Atlanta at , a quarter earlier.
Rent does not follow vacancy. Austin asks with vacancy and negative absorption of . Columbus asks , down from , with the tightest market of the five. A rent is a reading of what is listed, and the listings change from quarter to quarter.
Q3 2026. The source gives 88.2 % occupancy, which we turn into 11.8 % vacancy. Rent is net, $26.33, up 2.7 %. Space under construction is 826,725 sq. ft., down 33 %. The same source gives two different absorption figures for the year, so we show none.
Each row is a different publisher and period, and the definitions differ (direct against total, net against triple-net rent). Read the shape, do not rank to the decimal. A bar is missing where the report gave no figure.
This is five metros we could open, not a ranking of the country. We left out metros where the report gave a vacancy figure without a period or a method.
Vacancy is half the picture. The other half is what buyers pay and what gets built.
Buyers pay between 6.2 % and 7.3 %, and completions have fallen by more than a fifth.
The brokerage’s cap rate average is , and a broker survey puts the national average at . The split is by buyer: health systems at , provider owners at . Sales volume is up: in the quarter and over four quarters, up . The broker survey counts against a year earlier.
On supply, trailing-year construction starts are against a year before, down , and completions fell from to , down . Inventory is growing about a year, against a year in 2016 to 2022. In the quarter, was finished, against just over a year earlier. For office the same brokerage reports of completions, down . Both types are building almost nothing.
One metro, one quarter: 7.63 %. The report gave no price per foot.
Cap rates from different publishers and buyer groups are not one series. The two views show what each source measured.
A thin pipeline lifts the rent of the buildings that exist, as long as demand holds. It does not make a given building a good one. A building on a health-system campus, a building off campus with a leasing gap and an ambulatory center with one tenant are three different credits.
Price is what the buyer will pay today. Debt is what the building owes tomorrow.
Nobody we could read publishes a maturity schedule for medical office.
We looked for the amount of medical office debt coming due in 2026 to 2028 and did not find it. The securitized-loan delinquency series we found reports for office in January 2026 and does not track medical office on its own, so that number should not be used as a benchmark for the type. The second-hand term sheets we found put bank debt at and to of value, life companies at and to , and securitized lenders at with debt-yield floors of to .
A loan written at a coupon and refinanced at those terms carries a constant of , which is about the brokerage’s cap rate of . A loan sized on debt service at that constant gets less proceeds than one sized at a lower constant, if income is flat. That is our arithmetic on a model loan, not an observed maturity.
What this page does not replace: a rent roll and lease abstract for the building, a tenant credit review, an appraisal, a property condition report or a lender’s own loan file. A market reading says nothing about one loan.
Set the market against the building in front of you.
What to do with this
If you lend against medical office or are buying it: price the building from its own leases, not from the national average, ask who the tenants are and when their leases end, and treat the second-hand financing terms as a question for your lender.
Titleman reads the offering memorandum and the operating statement, rebuilds the numbers, and returns the issues and the questions to put to the seller, including what the rent roll says about lease terms and tenant concentration.
Sources
Found: from a dated document. Derived: calculated by us from found figures. Estimate: our judgment, with the method.
- FoundU.S. medical outpatient buildings, Q3 2026. CBRE, Q3 2026 U.S. Medical Outpatient Buildings Market Report, 5 Nov 2026. Read 19 Nov 2026. document ↗
- FoundU.S. office, Q3 2026. CBRE, Q3 2026 U.S. Office Market Report, 23 Oct 2026. Read 19 Nov 2026. document ↗
- FoundMedical office occupancy, cap rates by buyer and supply, Q3 2026. Coy Davidson, What healthcare occupiers should know about the MOB market, 31 Oct 2026, citing a research-firm dataset. Read 19 Nov 2026. document ↗
- FoundHouston medical office, Q3 2026. Coy Davidson, 31 Oct 2026. The source conflicts with itself on absorption and deliveries; we use only the figures that agree. Read 19 Nov 2026. document ↗
- FoundAustin and San Antonio medical office, Q3 2026. Matthews, Q3 2026 market reports, 30 Oct and 29 Oct 2026. Read 19 Nov 2026. document ↗
- FoundColumbus medical office, Q3 2026. Colliers, Greater Columbus Region, Q3 2026. Read 19 Nov 2026. document ↗
- FoundAtlanta medical office, Q2 2026. Lee & Associates Atlanta, Q2 2026 Atlanta Medical Office Market Report, July 2026. One quarter older than the other metros. Read 19 Nov 2026. document ↗
- FoundMedical office debt terms and loan delinquency, as reported in 2026. MMCG Invest, U.S. Medical Office Market Outlook 2026, which cites a loan-data research firm and lender term sheets. A second-hand reading; we did not open the term sheets. Read 19 Nov 2026. document ↗
- DerivedMedical outpatient against office: vacancy and rent gap. Titleman calculation from cards F1 and F2, the same brokerage and quarter.
- DerivedChange in construction starts and completions. Titleman calculation from card F3. Both figures are trailing-year and from one source.
- DerivedLoan constant of a model loan. Titleman calculation using the constant on card F8 and the cap rate on card F1.
- EstimateHow we handled the readings, and what we left out. Titleman estimate of the method. Ask us for the working: alexgalt@titleman.ai.
Figures are as reported in brokerage and research publications for the third quarter of 2026 (Atlanta: second quarter, July 2026), and an undated trade article on medical office financing, all read on 19 November 2026. Shares and sums marked as derived are Titleman calculations. Titleman’s opinion, not investment, appraisal or legal advice.