In June 2026 the USMLE programme announced that it would stop offering Step exams on a rolling year-round basis and move to a limited set of designated testing dates. In August it revised the timeline: the change now takes effect in January 2029 rather than 2028.
This is the largest structural change to how the exams are scheduled in years, and it is worth separating what has actually been announced from what people are guessing.
What has been confirmed
The model reduces the number of Step exam administrations each year and adds what the programme describes as safeguards intended to reduce security risks and promote fairness.
It takes effect in January 2029. The delay from the original 2028 date was attributed to feedback from engagement sessions with the medical education community, and to giving schools time to review and adjust policies and curricula.
An implementation panel was due to be established by the end of August 2026, and the programme committed to publishing further detail on applying, scheduling and exam day logistics as the work progresses.
What has not
The number of dates. This is the question everyone wants answered and it has not been published. The programme indicated it would share a resource on appointment availability drawn from historical testing volumes, which suggests capacity is being sized against current demand rather than reduced arbitrarily.
Whether the dates differ by Step. Step 1, Step 2 CK and Step 3 sit at very different points in a career and have different volumes, so a single calendar for all three would be surprising. Nothing has been said either way.
How rescheduling will work. Under year-round testing, moving an exam is inconvenient. Under fixed dates it could mean waiting for the next window, which changes the calculation entirely.
Why the reasoning matters
The stated rationale is security and fairness rather than capacity or cost. Under continuous testing, examinees who sit later in a cycle can in principle benefit from information circulating about recent forms. Concentrating administrations narrows that window.
That framing tells you something about what the final model will look like. If the goal were administrative, you would expect a small reduction in dates. If the goal is limiting information flow between test-takers, the dates will be genuinely concentrated.
What to do now
If you are testing before 2029, nothing changes. Do not restructure a study plan around a change that has not arrived.
If you are early in medical school and will be testing in 2029 or later, the one thing worth internalising now is that flexibility is going to be worth less than it used to be. The habit of leaving the exam date loose and deciding when you feel ready will not survive a fixed calendar. Planning backwards from a known date is a skill worth building before it becomes compulsory.
Whichever question bank you settle on, the part that matters under fixed dates is whether it tells you where you actually stand with enough lead time to act. QBankly reports by system and discipline rather than as a single percentage, which is the sort of signal that becomes more valuable when the date cannot move.
Watch for
The appointment availability resource, which will answer the capacity question. The implementation panel’s output. And any indication of how the designated dates interact with the application calendar, which is where the practical consequences for Step 2 CK will land.
Frequently asked questions
Does this affect exams taken before 2029?
No. Scheduling continues as it is today until the new model takes effect, so anyone testing in 2026, 2027 or 2028 is unaffected.
How many dates will there be?
Not published yet. The programme said it would release information on appointment availability based on historical testing volumes, and an implementation panel was being formed as of August 2026.