Stepora Blog · UWorld

How to Review UWorld Questions: The 5-Bucket Method That Turns Misses Into Tomorrow's Drills

Published 28 October 2026 · Stepora

You finished the block. Now comes the part that actually moves your score — and the part most students skip or rush. Reviewing a UWorld block is not re-reading explanations until your eyes glaze over. It is a structured pass with one job: for every question you missed, name exactly why you missed it, and turn that reason into something you will do tomorrow.

First, budget the review — or it will not happen

A timed 40-question block takes roughly an hour to sit through, and 1.5–2.5 hours to review properly. That makes one honest block a 2.5–3.5-hour commitment, not a one-hour task. Most review failures are really scheduling failures: a student plans four blocks a day, finishes the doing, and discovers at midnight that the reviewing never fit.

So the first rule of reviewing is set before you open the first question: never schedule more blocks than you can fully review the same day. In dedicated, that usually means 2–3 blocks (80–120 questions) with full review. A reviewed 40 beats two skimmed 40s every time — in the Deng et al. Step 2 CK data, every ~445 additional practice questions tracked with about one more point, but only because those questions were studied, not merely completed.

The 5-bucket error log

Open your block, explanations on, and go question by question. For every miss, write one line in a running log — the question's topic, and exactly one bucket:

One bucket per miss, no hedging. A miss tagged "kind of a knowledge gap, kind of a misread" is a miss you have not diagnosed, and an undiagnosed miss gets the wrong fix. If you genuinely cannot tell, default to knowledge gap and check the explanation again — the explanation usually tells you which it was.

The review pass, step by step

1. Work in the order you answered, not in score order. Reviewing in the order you sat the block keeps the reasoning fresh — you remember what you were thinking, which is the whole point of diagnosing a misread or a second-guess.
2. Start with the misses, then skim the rights. Misses get the full treatment: explanation, one-sentence rule, bucket tag, log line. Questions you got right get a fast pass — one line on why the right answer was right, and a check that you got it right for the right reason. A lucky guess is a miss wearing a costume; log it as one.
3. Write the rule, don't just read it. For every miss, write one sentence: the rule the question tested, in your own words. Ten written rules at the end of a review session beat ten read explanations — reading feels like learning, writing proves it.
4. Close the loop in the log. End the session by counting the buckets. Whichever bucket holds the most misses owns tomorrow's drill time — that is the entire point of the log. A review session that ends without a plan for tomorrow was a reading session, not a review session.

Turning misses into tomorrow's drills

This is where the 5-bucket log earns its keep. Your biggest bucket is not a statistic — it is tomorrow's first task:

• Knowledge gaps clustered in one system: tomorrow starts with a targeted re-learn on that system, then a drill set drawn from your incorrects on it. Not "study more" — study that.
• Misreads or second-guesses leading: tomorrow's block is done with the strict reading protocol — question first, decisive detail marked, answer changed only on new evidence. The drill is the method, not the content.
• Reasoning errors leading: tomorrow's review goes deeper, not longer — every miss becomes a written rule, and the rules get re-read before the next block.
• Timing leading: tomorrow's block is a pacing drill with checkpoints, and the last ten questions get the same calm as the first ten.

Over a week, the bucket counts tell you what no single percentage can: whether your prep has a content problem, a reading problem, a confidence problem, or a pacing problem. And when a plateau arrives — two NBME forms returning the same score — the same buckets are the diagnosis.

What not to do

Don't carry review debt. Done questions piling up unreviewed is the single most common way a prep quietly stalls: the question count rises while the NBME trend stays flat. If debt has already built up, stop doing new blocks until it is cleared — review debt is paid by doing less, not by working longer.

Don't review only your strengths. It feels productive to re-read explanations for topics you already know. It is the review equivalent of re-doing questions you already got right — comfortable, and useless. Your log decides the drills; your feelings do not.

Don't let the review eat the NBME chain. Reviews are daily work; NBMEs are the decision-grade assessment. Keep the cadence — current online forms 28–34 at $62 each, 7–14 days apart, last full form at least 7 days before the exam, Free 120 3–7 days out — and let the NBMEs, not your block percentages, decide the exam date. Two recent forms at about 69–70% correct (EPC) is the green light.

Done right, a review session is the most honest hour in your prep: it tells you, in writing, exactly what is still broken and exactly what tomorrow fixes first. Fit it into a full day with the Step 1 study schedule, or let the schedule maker build the review time into your day automatically — and if your block percentages are confusing you, here is what they actually mean.