← News and Insights

By Golden Triangle Consultants

US MFN Drug Pricing: What Changed Between 2025 and 2026?

US MFN pricing has moved from political ambition to deals and draft rules. What has changed, why Medicare timelines look uncertain, and what to model next.

When we wrote about Most-Favoured-Nation (MFN) pricing in May 2025, much of the discussion concerned how an ambitious executive order might become policy. Sixteen months later, manufacturers have signed pricing agreements with the US government, some linked to tariff relief, and detailed Medicare rules have been proposed. Yet companies still do not have a settled set of MFN rules to plan around.

MFN modelling is now viable and necessary to inform pricing decisions. The emerging detail gives teams a basis for assessing potential impacts, but major questions remain over which prices will count, which products will be affected and when obligations will begin. Modelling should therefore compare a range of plausible outcomes and show how exposure changes under different assumptions.

From a price target to a series of deals

The starting position was broad. In May 2025, HHS set a target based on the lowest price in an OECD country with GDP per capita at least 60% of the US level. The expectation covered branded products without generic or biosimilar competition. It was a statement of what the administration wanted manufacturers to deliver, rather than a complete operating model.

The next phase centred on agreements with individual manufacturers. Pfizer's September 2025 agreement linked pricing commitments with investment and a conditional three-year tariff grace period. Some agreement terms remain confidential, so public announcements alone cannot establish a company's precise obligations. By 31 August 2026, the White House reported agreements with 26 manufacturers. The announcements covered Medicaid pricing, new medicines and direct purchasing, with TrumpRx launching in February 2026.

CMS also developed GENEROUS, a voluntary Medicaid model using negotiated supplemental rebates to bring participating products' net prices into line with selected international prices. Participation and product coverage matter: an announcement of an MFN agreement does not establish the same exposure across every medicine and US payer.

Even the meaning of “MFN price” has evolved

The Council of Economic Advisers' May 2026 account of the voluntary agreements describes a more specific benchmark: the second-lowest adjusted net price across the other G7 countries, Denmark and Switzerland. It includes rebates and other concessions, including portfolio-wide clawbacks, and adjusts international prices using GDP per capita at purchasing power parity.

This differs from the initial lowest-price target across a broader group of OECD countries. The voluntary-agreement benchmark also cannot simply be substituted for the separate calculations proposed for Medicare. A model built around the original headline needs to be revisited as the mechanism becomes clearer.

Under the UK–US pharmaceutical arrangement published in April 2026, the UK committed to increase spending on new medicines and net prices for future launches. The US, in turn, agreed that a new medicine's UK price would not set its Medicaid MFN price when it was the lowest in the basket, consistent with GENEROUS. Global pricing teams therefore need to follow negotiations outside the US as well as US reimbursement policy.

GLOBE and GUARD: detailed rules, uncertain start dates

GLOBE's Part B proposal and GUARD's Part D proposal would require manufacturers of eligible drugs to pay rebates calculated against international price benchmarks. Participation would be compulsory, with each model operating in selected geographic areas covering approximately a quarter of the relevant Medicare population. Their draft regulations set out eligibility, reference countries, pricing data and rebate calculations in detail. Neither model has been finalised.

This gives companies something concrete to model. For example, GLOBE proposes alternative benchmark calculations using existing international pricing data or manufacturer-submitted net-price data. Those choices can produce different results for the same product. The detail deserves attention even while the implementation timetable remains uncertain.

The proposed starts were 1 October 2026 for GLOBE and 1 January 2027 for GUARD. As of 18 September, the regulatory review records still show both GLOBE and GUARD as pending at the Office of Information and Regulatory Affairs. Their designation as being at the “final rule” stage refers to the review process; it does not mean a final rule has been published.

With less than a fortnight until GLOBE's proposed start, our assessment is that a delay looks increasingly likely. GUARD has more time, but its January date should also be treated as uncertain. We would plan for slippage and possible redesign rather than assume implementation exactly as drafted.

What should a pricing team do with that uncertainty?

Our view is that MFN is now too embedded in US pricing policy to plan on it disappearing. A delay to the Medicare models would give companies more preparation time while manufacturer agreements, Medicaid arrangements and international negotiating pressure continued to shape decisions.

A useful forecast should start with the obligations already applicable to the company, then compare several credible outcomes:

  • The Medicare proposals proceed broadly as written, on the proposed timetable or with later start dates.
  • Final rules change the eligible products, benchmark methods or other material terms.
  • Compulsory Medicare models stall for longer, while negotiated MFN arrangements continue to develop.

Within each outcome, the important sensitivities include reference-country prices, net-price assumptions, currency and purchasing-power adjustments, product scope and timing. The analysis should show how these interact with existing International Reference Pricing rules and with commercial decisions in individual markets.

A local price concession might have a substantial US consequence under one mechanism and a much smaller effect under another. A delayed launch may protect a benchmark temporarily while sacrificing local revenue and patient access. These are decisions to compare explicitly, using ranges where the data or policy is still unresolved.

For a pricing team, the practical question is which pricing decisions would still make sense if the rules or start dates changed. GTC's IRP, MFN and global pricing reviews bring these questions into product and portfolio planning.

Policy position reviewed on 18 September 2026. Expectations about delays and future outcomes are GTC's assessment of the sources available on that date.

Discuss your global pricing challenge

Explore the implications for your product, launch plan or portfolio.

Talk to GTC