How Many UWorld Questions Per Day? Build a Pace You Can Sustain

The right number of UWorld questions is the number you can answer, review, and learn from within the time you actually have. Start with your review capacity, then set the daily target. A larger completed-question count is useful only if it leaves you better prepared for the next unfamiliar question.

It is easy to borrow someone else’s number. Forty sounds manageable. Eighty sounds serious. A hundred and sixty sounds like proof that you are working hard enough. None of those numbers tells you whether the person posting it had your baseline, responsibilities, or review process.

What research supports, and what it cannot tell you

In a single-school study, Burk-Rafel and colleagues surveyed 274 students after scored Step 1 in 2014 and 2015. Completing more practice questions was associated with higher scores after adjustment for other measured factors. The study did not randomize students to daily quotas, identify an optimal number of UWorld questions, or establish the same effect under today’s pass/fail exam. Its reported model fit describes multiple predictors together, not question volume alone. Burk-Rafel et al., 2017

The learning case for questions is broader than that association. In laboratory experiments, retrieval practice improved delayed retention compared with repeated study. Continued retrieval after an initial correct response also benefited later recall. These studies support trying to retrieve an answer and revisiting learning; they do not establish a universal fatigue threshold or a mandatory review-time ratio. Roediger and Karpicke and Karpicke and Roediger

Calculate a starting pace from a real study day

For two or three ordinary days, record how long a small question set takes, how long review takes, and what remains unfinished. Include correct answers you guessed and explanations you could not confidently teach back. Do not measure only your fastest day.

Daily question capacity = minutes available for questions and review, divided by your observed average minutes per question including review.

Imagine you have three hours for question work after allowing time for meals, breaks, classes, and targeted content study. If answering and reviewing one question averages six minutes, 180 divided by six gives a starting target of 30 questions. If your average later falls to four minutes while review remains effective, the same time budget accommodates about 45. These are invented planning examples, not recommended quotas.

Leave room for unfamiliar material. An endocrine set may take longer than familiar cardiology. A useful plan has enough flexibility for that difference instead of making every slower review feel like a failure.

Review the decision, not just the correct option

After a miss or uncertain correct answer, write the shortest explanation that would help with a new version of the problem: what was the question asking you to decide; which finding supported the answer; why was the closest alternative less appropriate; and was the gap factual knowledge, interpretation, or execution? Then decide what you will retrieve or practice later to check that the gap closed.

“Review renal” is too broad to guide tomorrow’s work. “I confused increased hormone production with hormone release; revisit the discriminator and test it on a new problem” gives a specific task without copying an explanation.

Choose question mode for the task

Use targeted sets when working on an identified gap. Use mixed sets to practice recognizing what an unfamiliar question is testing. Timed practice helps you see whether your process fits the available time; untimed review gives room to repair it. There is no validated universal percentage required before using timed mode.

For Step 1 examinations on or after May 14, 2026, USMLE uses fourteen 30-minute blocks with up to 20 questions per block across an eight-hour testing session. Practice with the current official interface. A 40-question Qbank set can still be useful, but it is not one current Step 1 exam block. USMLE software update

Adjust once a week

Look at completed review, performance on unfamiliar material, and whether the schedule feels sustainable. If unanswered review accumulates, reduce new-question volume. If the same gap returns, add focused retrieval and another unfamiliar example. If review is complete and sustainable, try a modest increase and reassess.

Do not interpret a rising Qbank percentage as a guaranteed Step 1 result. Subject mix, repeated questions, timing, and references affect that number. Use official assessments and their reporting guidance for readiness decisions. See the NBME interpretation guide, UWorld and AMBOSS guide, and study-schedule guide. Book a free study-planning consult if you need help organizing your available time and practice data.