This audit is for family medicine candidates deciding whether UWorld's ABFM QBank should be their primary bank for the one-day certification exam. It addresses the applied-reasoning layer — high-fidelity single-best-answer practice with deep explanations. The principal limitation is not quality but finiteness and recognition: a fixed bank, worked twice, starts measuring how well you recognise its own items rather than how you perform on unseen ones, and its headline total tells you nothing about balance across the five domains.
What UWorld offers for ABFM right now
Figures are vendor-reported from medical.uworld.com and were last checked on 20 July 2026; confirm current inclusions and pricing on the product page.
| Item | Vendor-reported detail (20 July 2026) |
|---|---|
| Question count | "1,150+" practice questions, at or above exam-level difficulty |
| Price (QBank) | About US$499 (90 days) to US$749 (730 days); medical-library add-on about US$50 more per tier |
| Self-assessment | Three forms, about US$50 each or US$120 for all three; each 4 blocks of 40 questions |
| Learning tools | Spaced-repetition flashcards, notebook, study planner |
| Analytics | Performance tracking by subject, system and date |
| Library | Optional "1,300+" peer-reviewed articles |
| Adaptive / AI | Active-learning tools and analytics; no adaptive difficulty engine advertised |
Note the last row: UWorld's flashcards use spaced repetition, but the QBank itself is a fixed, expertly authored set, not an adaptive engine. That is a strength for fidelity and a limitation for measurement, and both matter below.
The ABFM exam anchor
The one-day Family Medicine Certification Examination is 300 single-best-answer questions across four 75-question sections of 95 minutes, around 100 minutes of pooled break time, roughly six hours and twenty minutes in total, at Prometric, with no return to a closed section. The 2025 blueprint uses five domains of care defined by clinical activity: Acute Care and Diagnosis (about 35%), Chronic Care Management (about 25%), Emergent and Urgent Care (about 20%), Preventive Care (about 15%) and Foundations of Care (about 5%). Verify current weightings on theabfm.org.
Reading the live count by blueprint domain
Do not take "1,150+" as a coverage figure. UWorld organises its bank by subject and organ system, while the exam is organised by clinical activity, so a headline total can hide a thin domain. The audit step is to run your recent UWorld sessions through the five-domain lens yourself: tag each question by domain of care and count how many you have actually seen in Emergent and Urgent Care and in Preventive Care specifically, since those cut across systems. A bank can be large in aggregate and still leave a cross-cutting domain under-practised. A worked example makes the risk concrete: suppose your analytics show 900 questions attempted, and you tag a recent 100-question sample by domain — 42 fall under Acute Care and Diagnosis, 30 under Chronic Care Management, 14 under Emergent and Urgent Care, 10 under Preventive Care and 4 under Foundations of Care. That looks defensible against the blueprint until you check your accuracy and find the two smallest buckets are also your weakest, and those are precisely the buckets a headline total told you nothing about. The method for building this map is in completion is not coverage.
Sample the question style
UWorld's house style is well suited to the family medicine exam, and this is where most of the transfer happens.
- Recall versus application: heavily application-weighted — questions ask for the next best step or the best management, not a bare fact. This matches the exam's reasoning demand.
- Stem length: long, detail-rich vignettes. Good preparation for the exam's reading load, though the exam's stems are often shorter, so do not over-train on marathon stems alone.
- Option plausibility: distractors are close and designed to catch specific errors, which trains genuine discrimination rather than elimination-by-absurdity.
- Image and data interpretation: present and useful, matching the exam's data-interpretation items.
- Management sequencing: strong emphasis on "what next" reasoning, which transfers directly.
The style transfers well. What transfers less well is anything time-sensitive: a fixed explanation written months ago can lag a changed guideline.
Jurisdiction and recency
The exam expects current US practice. Take a stratified sample across domains and check each item's explanation against current primary guidance — USPSTF for screening, ACIP for immunisation, ADA for diabetes targets, and the relevant cardiology guidance for blood-pressure and lipid thresholds — and record the review date. UWorld updates content, but no static explanation is guaranteed current on the day you read it. If a screening-age or vaccine-schedule recommendation has shifted since the explanation was written, the bank can teach you a now-outdated interval with complete confidence — which is why the primary source, not the explanation's tone, decides the answer. Any preventive-interval or treatment-target item is a recency checkpoint, not a fact to memorise blind.
The format gap
Be clear about what a standard question bank can and cannot do. It can train diagnostic and management reasoning to a high level. It is weaker on three things the exam still tests: preventive-care intervals, which depend on current schedules rather than reasoning; ambulatory longitudinal care, the management of a comorbid patient over time, which a single-encounter vignette only approximates; and US guideline updates, where static content carries recency risk. Cover these by verifying against primary guidance and by testing on freshly written unseen items rather than assuming the bank has done it for you.
Duplication and contamination
The central risk with any finite bank is that completion becomes recognition. On a second pass you increasingly remember the answer rather than reason to it, and near-duplicate stems on the same concept inflate your percentage without adding capability. Two defences: first, do not treat a rising second-pass percentage as rising readiness — measure on unseen items instead; second, if you add a second bank, follow the two-Q-bank rule so the second bank stays unseen and measures transfer rather than duplicating practice. The wider point about misreading a bank percentage is in why your Q-bank percentage is not your exam score.
Best-fit matrix
Decide what job this bank is doing for you.
| Use case | Fit | Note |
|---|---|---|
| Foundation building | Moderate | Strong explanations teach, but a weak-foundation learner may need video first |
| First full pass | Strong | Application-heavy style transfers well |
| Second bank | Moderate | Only if kept unseen and used for measurement |
| Retake | Strong | Deep explanations diagnose reasoning errors |
| Final simulation | Partial | Self-assessment forms help; still verify five-domain balance |
A seven-day plan for family physicians
No proprietary-algorithm claims — this works with any bank.
- Monday: One timed UWorld block in your weakest domain; commit answers before reviewing.
- Tuesday: Review only wrong and flagged items; turn each into a spaced-repetition flashcard.
- Wednesday: Verify every guideline-dependent item from Monday against current US guidance; note review dates.
- Thursday: A fresh, timed, unseen ABFM block in iatroX — because these items are outside UWorld, the score measures transfer, not memory.
- Friday: Re-test Tuesday's flashcards; log persistent misses by domain.
- Weekend: One UWorld self-assessment form for a pacing read; move unresolved topics to next Monday.
UWorld does one job — high-fidelity applied practice with deep explanations. iatroX does a different job — unseen measurement. Kept separate, they do not duplicate.
Decision checklist: continue, supplement, switch or stop
- Continue while first-pass unseen performance is rising and your five-domain profile is levelling up.
- Supplement with an unseen measurement bank once you near completion, so a second pass does not masquerade as progress.
- Switch or de-prioritise only if a whole domain is under-represented for your needs and cannot be filled — rare with UWorld, but check Preventive Care and Emergent and Urgent Care specifically.
- Stop new questions and move to full-length timed papers when errors are pacing and technique rather than knowledge.
Three mistakes this audit is designed to stop
First, reading the headline count as coverage: "over a thousand questions" feels comprehensive, but if only a handful touch a cross-cutting domain such as Preventive Care you are under-prepared in an area worth roughly one mark in seven, so count by domain rather than in aggregate. Second, trusting a rising second-pass percentage: the number climbs because you remember the items, not because you have improved, and the only honest readiness figure comes from unseen material. Third, assuming the explanations are current: they are written to a date, while screening intervals, immunisation schedules and disease targets move, so a confident explanation is never a substitute for checking the primary source on anything time-sensitive.
Bottom line
UWorld's ABFM QBank has a question style that transfers well to the real exam — application-weighted, close distractors, data interpretation, next-step reasoning. What does not transfer is anything the format cannot carry: preventive-care currency, true longitudinal management and, on a second pass, honest measurement, because recognition creeps in. Use it as a strong primary bank, verify recency against primary guidance, map your coverage onto the five domains, and keep an unseen source for readiness.
Frequently asked questions
Is UWorld ABFM enough for ABFM on its own? For many candidates it is a sufficient primary bank because its style closely matches the exam's reasoning demand, but sufficiency should be judged on unseen, timed performance across the five domains, not on finishing the bank. Its main risk is recognition on a second pass, so pair it with a small unseen sample late in preparation to keep your readiness signal honest.
Which ABFM component does UWorld ABFM not reproduce well? Preventive-care intervals and longitudinal ambulatory management, plus recency on any recently changed guideline. A single-encounter vignette approximates but does not fully reproduce managing a comorbid patient over time, and static explanations carry currency risk. Verify screening and immunisation items against USPSTF and ACIP rather than trusting an explanation's date.
How many UWorld ABFM questions should I complete per day for ABFM? Tie it to your timeline rather than a fixed quota. Over eight to twelve weeks, one to two timed blocks a day with same-day review is sustainable and lets the spaced-repetition flashcards do their work. Volume without review, or review without spacing, wastes the bank's main strengths.
When should I stop using UWorld ABFM and move to mixed mocks? When your errors are dominated by pacing, stamina and misreading long stems rather than missing knowledge, and when a second pass is producing recognition rather than reasoning. Full-length, timed, mixed papers then train exam execution across four 95-minute sections better than another fresh block.
How should I combine UWorld ABFM with iatroX without duplicating practice? Keep the jobs distinct: UWorld for applied practice and explanation, iatroX for unseen transfer measurement on items you have not seen. Never review the same question in both, and treat the iatroX score — not your UWorld percentage — as your readiness read. The two-bank logic is set out in the two-Q-bank rule.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026. Question counts, prices, self-assessment details and library inclusions are vendor-reported from medical.uworld.com on that date and change over time; confirm them on the product page before purchase. Disclosure: iatroX operates a competing question bank; its role in this audit is confined to unseen transfer measurement — a job UWorld's fixed bank does not claim — and this article does not disparage UWorld's authoring quality, which is a genuine strength. Corrections are welcome via the feedback route on iatrox.com.
References: American Board of Family Medicine — exam blueprint and one-day exam pages (theabfm.org); UWorld ABFM product page (medical.uworld.com/abfm); USPSTF and ACIP for preventive-care currency; iatroX internal resources — completion is not coverage and why your Q-bank percentage is not your exam score.