In October, Berkeley Research Group published “The Size of the 340B Program.” It estimates that payers and patients spent $244.3 billion in 2025 on drugs bought through 340B, that covered entities and their partners kept a $142.2 billion margin, and that 340B drugs made up 26.5% of net U.S. drug spending. Its last page discloses PhRMA funding.

Before you repeat those figures, or rebut them, know where they come from. BRG starts with HRSA’s $100.0 billion in reported 340B purchases and builds the $244.3 billion through a chain of assumptions about discounts, payer mix and reimbursement. The report publishes point estimates with no sensitivity ranges and too little detail to reproduce each step.

Industry funding does not make the work wrong. Hospital-funded studies deserve the same scrutiny. Some BRG choices push its estimate down. It assumes the prime vendor’s data capture 98% of 340B sales, above the 90% to 95% range MedPAC reported in 2015, and it assumes covered entities earn nothing on cash prescriptions. The right test is whether the method holds up. Three assumptions carry most of the weight, and the report tests none of them.

Start with retail prices. BRG assumes payers reimburse every retail 340B prescription at wholesale acquisition cost, the list price, except Medicaid fee-for-service and cash claims, which it sets at cost. Wholesale acquisition cost is a benchmark. The report cites no claims data showing that plans pay it, and it never shows how the total moves if they pay less.

The retail payer mix comes from one state: Minnesota’s 340B covered entity report from November 2024. BRG uses national CMS data for the non-retail split. It offers no test of whether Minnesota’s mix looks like the rest of the country.

For private non-retail drugs, BRG assumes payers reimburse 3.08 times the 340B price. That figure comes from a 2024 New England Journal of Medicine study by James Robinson, Christopher Whaley and Sanket Dhruva. The study used 2020 and 2021 Blue Cross Blue Shield claims for 404,443 patients on infused drugs. It observed what the plans paid and modeled what hospitals paid manufacturers. Whether it fits all private non-retail 340B spending in 2025 is a separate question the report does not answer.

The report also skips sensitivity analysis. BRG publishes no scenarios and no uncertainty bands. A reader has no way to see how far $244.3 billion shifts under a different retail price, payer mix, private multiple or discount estimate. That is the report’s largest gap.

Two framing errors will travel with the number, though BRG commits neither. The first treats margin as profit. BRG’s $142.2 billion equals modeled reimbursement minus acquisition cost, shared among covered entities, contract pharmacies, technology vendors and consultants. Hospital net income comes only after the costs of running the program. The second treats total spending on 340B drugs as the program’s cost. BRG estimates what payers spent on drugs bought at the 340B price. It does not estimate what they would have spent without 340B. That requires a counterfactual on prices, drug choice, volume and site of care, and BRG does not attempt one.

The report gets several things right. The program is large and growing. BRG cites the Congressional Budget Office’s finding that hospital integration with off-site clinics, Affordable Care Act-era participation growth and contract pharmacy use drove most of that growth. Its call for public data on 340B margins and how covered entities use them is fair. Hospitals that respond to BRG with outrage instead of numbers concede that point by default.

Cite the estimate with its assumptions. If you use $244.3 billion, say in the same sentence that it is a modeled figure that prices most retail claims at list and draws its retail payer mix from one state.

Ask BRG for the sensitivity runs. Request scenarios that change one input at a time, with the evidence for each alternative.

Answer with your own ledger. A covered entity that shows its actual reimbursement by payer, its acquisition cost and where the margin went holds a stronger position than one that disputes a national model in the abstract.

The report documents a big program. It does not settle how big, or what it costs. Policymakers should ask how much the answer changes when the assumptions do.