This workflow is for family medicine candidates using the UWorld ABFM question bank for the written certification exam. It addresses the knowledge base across all five domains of care. The principal limitation is inherent to any finite premium bank: with around 1,150 questions, a second or third pass rewards memory of the specific item rather than mastery of the principle, so your repeat accuracy can look excellent while your performance on genuinely new questions stalls.
What UWorld ABFM offers right now
The figures below are vendor-reported and last checked 20 July 2026. Prices and counts change, so confirm them on the product page before you subscribe.
| Attribute | What UWorld reports (vendor-reported, 20 July 2026) |
|---|---|
| Bank size | 1,150+ questions at or above exam-level difficulty, covering the ABFM blueprint |
| Access and price | 90 days $499, 180 days $549, 360 days $599, 730 days $749; add roughly $50 per tier for the Medical Library |
| Self-assessments | Forms 1 to 3 at $50 each or $120 bundled; each form is 4 blocks of 40 questions |
| Extras | Biostatistics Review at $25 |
| Features | UAsk AI study assistant, spaced-repetition flashcards, visual explanations with peer-selection percentages, a digital notebook, a study planner and performance analytics by subject, system and date |
| Apps | Native iOS and Android with cross-device sync |
Two honest notes. First, UWorld ABFM is a fixed bank you drive through blocks in tutor or timed mode; the UAsk assistant helps you interrogate explanations and the flashcards use spaced repetition, but the question selection is yours, not a proprietary adaptive engine that guarantees coverage. Second, at around 1,150 questions this is a curated, high-quality bank rather than a high-volume one, which makes the memorisation risk more acute and the discipline in this article more important.
The exam you are actually sitting: the ABFM one-day format
The one-day exam is 300 single best answer questions in four sections of 75 questions, 95 minutes per section, with about 100 minutes of poolable break time and roughly six hours and twenty minutes overall at Prometric. Navigation is forward and back within a section, but a submitted section cannot be reopened. Since 2025 the blueprint is five domains of care based on clinical activities: Acute Care and Diagnosis around 35 per cent, Chronic Care Management around 25 per cent, Emergent and Urgent Care around 20 per cent, Preventive Care around 15 per cent, and Foundations of Care around 5 per cent. Confirm the current weightings on theabfm.org, and use them, not organ systems, to balance your practice.
Build a blueprint inventory and protect an unseen pool
Before you start grinding blocks, lay the bank out against the five domains and record how many questions sit in each. Then, deliberately, reserve a protected pool of questions you will not touch during learning, so you have unseen material for timed mixed assessment later. If you consume the whole bank in tutor mode first, you have no clean way to measure readiness, because every later block is a memory test. The protected pool is what keeps a finite bank honest.
First pass: mixed by default, filtered only where foundations are weak
Start mixed. Topic-filtered blocks quietly remove the hardest step, which is deciding what kind of problem you are facing, so they flatter your accuracy and hide it from the real task. Use topic filters only where your foundations are genuinely weak and you need to build before you can test. Everywhere else, mixed blocks from the outset train the discrimination the exam actually assesses, and they stop the topic label from doing your thinking for you.
Review each miss with an error code and one action
When you get an item wrong, do not transcribe the whole explanation into your notes; that is copying, not learning. Assign one error code, knowledge gap, misread stem, premature closure, guideline lag, calculation slip or time pressure, and write one corrective action. One code, one action, one line. The discipline forces you to name why you missed it and to commit to a single next step, which is far more useful than a beautifully highlighted explanation you will never reread.
Use transfer practice before repeating an item
The single most effective habit for "not memorising the bank" is transfer practice: before you ever re-answer an original item, answer a new question that tests the same underlying principle. If you can get the fresh item right, you have the concept and can leave the original alone. If you cannot, you have found a real gap rather than a memory lapse. This is where a second, genuinely unseen bank earns its place, because UWorld's own items become familiar the moment you have seen them.
Switch to mixed timed blocks when domain floors are met
Move to mixed, timed blocks as soon as your domain floors are met, even if your first pass through the bank is incomplete. Finishing the bank is not the goal, and waiting for 100 per cent completion delays the timed practice that actually predicts performance. When every domain has been sampled to your floor and your accuracy is stable, the marginal value of one more tutor-mode block is lower than the value of a timed mixed assessment on unseen items.
Exit criteria that are not 100 per cent completion
Judge readiness on five things: coverage across all five domains, unseen timed performance that is stable at or above your target, comfortable pacing at roughly 75 seconds per item, retention that holds when a topic returns after a gap, and calibration against the official ABFM self-assessment. Notice that "finished the bank" is not on the list. A bank that is fully completed but not retained is a weaker signal than an unfinished bank whose unseen blocks you answer reliably.
A seven-day plan for a family physician
On day one, a mixed timed block in UWorld, every miss error-coded with one action, no transcription. On day two, the same, plus targeted flashcard review of the coded gaps. On day three, transfer practice in iatroX on the week's worst principles, so the check is on unseen items rather than UWorld replays. On day four, a light spaced review of earlier misses. On day five, a timed mixed block on unseen questions for a clean readiness measurement. On day six, a filtered block only in a domain where your floor is not yet met. On day seven, rest or brief consolidation. UWorld is the learning bank; iatroX is the unseen measurement bank. No proprietary-algorithm claims are made about either.
Decision checklist: continue, supplement, switch or stop
Continue if your unseen accuracy is rising and your domains are balancing out. Supplement with a second unseen bank and the official self-assessment once you have seen enough of UWorld that repeat accuracy is inflating and no longer informative. Switch only for a genuine content or currency problem, not for novelty. Stop first-pass learning and move to full-length timed simulation when your exit criteria are met. Every branch turns on a measurable gap, not on sunk cost or the urge to try something new.
The bottom line
UWorld ABFM is a strong, high-quality bank with excellent explanations, but it is finite, so the whole game is learning the principle rather than the item. Protect an unseen pool, start mixed, review by error code, use transfer practice before any repeat, and measure readiness on unseen timed blocks. Do that and the bank teaches you the exam; skip it and the bank quietly teaches you the bank.
Frequently asked questions
Is UWorld ABFM enough for ABFM on its own? It is a strong single bank, with around 1,150 high-quality questions on the vendor's figures, but its size makes memorisation a real risk on a second pass, so it is best paired with a second source of unseen questions and the official ABFM self-assessment. As a standalone it teaches well; as a measurement tool it becomes unreliable once you have seen the items.
Which ABFM component does UWorld ABFM not reproduce well? The full-length timed exam day. UWorld blocks are shorter than the exam's four 75-question sections, and no bank reproduces the roughly six-hour-twenty stamina and the no-going-back sectioning. Use the self-assessments and full-length timed simulation to build that endurance rather than relying on short tutor-mode blocks.
How many UWorld ABFM questions should I complete per day for ABFM? There is no official figure; this is guidance, not a vendor claim. One 40-question block, fully reviewed and error-coded, is a common and sustainable daily target, adjusted to your weeks-to-exam. The quality of your review matters more than the raw count, and a rushed second block you never analyse adds little.
When should I stop using UWorld ABFM and move to mixed mocks? Move to mixed timed mocks when your domain floors are met and your unseen mixed accuracy is stable, even if your first pass is incomplete. Completion is not the trigger; a stable performance signal on fresh, timed questions is, and chasing 100 per cent completion tends to delay the practice that actually predicts your result.
How should I combine UWorld ABFM with iatroX without duplicating practice? Apply the two-Q-bank rule: keep UWorld as your learning bank and use iatroX as your measurement bank on unseen items. The point is not to re-answer the same questions in both tools; you learn on UWorld and you test transfer on iatroX, which keeps the second bank genuinely unseen and your readiness signal clean.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026. All UWorld figures, features and prices are vendor-reported and may change; confirm them on the product page, and verify the ABFM blueprint weightings on theabfm.org. Disclosure: iatroX operates a competing question bank, so its role here is confined to a job UWorld does not need iatroX for, namely acting as a second, genuinely unseen bank for transfer measurement; the two are used side by side, not as substitutes. Corrections are welcome through the feedback route on iatrox.com.
References: ABFM Family Medicine Certification exam blueprint and 2025 redesign (theabfm.org); UWorld ABFM Family Medicine Board Review product page (medical.uworld.com/abfm); iatroX, "Your Q-Bank Percentage Is Not Your Exam Score" (iatrox.com/blog/qbank-percentage-not-your-exam-score); the two-Q-bank rule and blueprint-coverage-matrix pillars on iatrox.com; and the iatroX comparison hub (iatrox.com/compare).