The NBME retires forms. It does not retire logic. Forms 20–33 are already written down — 4,174 one-line recall cards, tutored, system by system.
The writers retire forms. They don’t retire logic. The concepts tested on forms 30–31 already appeared on 20–25. New vignette, same trap.
Which means the exam you are about to take has, in substance, already been published fourteen times.
This is the part students get wrong, so read it slowly.
Every NBME, forms 20 through 33, was gone through topic by topic. Every topic those forms tested was pulled out and written as a single one-liner — and then sorted into the system it belongs to.
That one-liner is not a summary. It is the thing the exam asks you. The line you are given is the line that comes back at you in a vignette.
And it is not just stated at you. Each one is taught — the concept explained underneath it, in plain language, so the point actually lodges instead of being read once and lost. You are not memorising a fact card. You are being tutored on the exact point, then asked it back.
That is the whole difference. The Foundation deck teaches you the book. The Booster teaches you the exam.
Students who run it properly — system-wise, to completion — have moved from 49% to 75% on their NBMEs in 15–20 days.
Not because they learned new medicine in three weeks. Because they stopped being surprised. Every pattern the last fourteen forms used was already in front of them, in one line, explained.
If you are on the 62–65% plateau, this is the third move on that page and the one marked most important. If you are trying to push a comfortable score into a strong one, it is the engine behind getting above 75%.
Take the form. Read the number. Feel something. Move on.
The one document that lists exactly what the exam wants from you — filed away, unread. Taking another form without harvesting the last one is paying for the same diagnosis twice.
Harvesting it is a skill in itself: the review protocol is how you work a form the same day you take it, and the MCQ protocol is how you read the question in the first place.
Every question you missed is the exam telling you, in writing, what it will ask again.
A form isn’t a verdict. It’s a reading list. The number on top is the least useful thing on the page — and if you are using that number to decide whether you are ready, read this instead.
| Order | System by system — never form by form |
| Rule | Finish a system completely before starting the next |
| Then | Attempt your next form and watch what stopped feeling new |
| Window | The final 1–2 weeks before your attempt |
Worked example. You took form 28 last week and form 29 is next week. Run the Booster across 20 through 28, every system, to completion. That is every point the exam has already asked, front-loaded into the days before you take the next one.
Which form to attempt and in what order is its own decision — that is the exact NBME order. If it is your first one, start with taking your first NBME.
This is a 1–2-weeks-out weapon, not a learning deck.
Building the book is a different job — that is the four-reading method and the six-day revision. Repairing a broken system is Mehlman, and question technique is the RPSR method for UWorld.
The Booster comes after all of that, and it sits inside the last two weeks — exactly where the 16-week plan puts it. Want a hard check before exam day? The full-length assessments give you a verdict.
| 14 | retired forms, NBME 20–33 |
| 4,174 | one-line cards — each one a topic the exam actually asked |
| Tutored | every line explained underneath, not just stated |
| System-wise | decks run to completion, one system at a time |
| Final 1–2 weeks | before the attempt — never earlier |
You don’t need another form. You need to stop missing what the last one already showed you.
Stuck at 62–65% — the three moves, with this deck as the third.
The exact NBME order — which form to attempt, and when.
The NBME review protocol — how to harvest a form the same day.
Foundation versus Booster — which deck does which job.
The 16-week plan — where the final two weeks actually sit.