Skip to main content

UWorld vs Alma Study for medical students: honest head to head guide

By , MD, FRCPC · Published · 24 min read

How we write and check articles: Editorial Policy.

A medical student at a desk choosing between a question bank path and a guided study path.

Disclosure: Alma publishes this article and makes one of the study tools it discusses.

If you are a medical student, we would not replace UWorld for Step 1 or Step 2 CK practice. Their Step 1 QBank page describes their prep as “the gold standard for mastering preclinical concepts and passing the exam.”[1] Alma Study sits beside it as a personalized tutor: it looks at your flashcards and quizzes on your own material, tells you what to study next, how to review it, and builds a day by day plan around your exam date.

If you can afford both, the most effective setup for many students is: UWorld for high quality questions and self assessments, Alma to turn what you miss and what you learn in class into a spaced, scheduled, data driven study system. If you cannot buy both, we outline below when to pick UWorld only, Alma only, or both together.

Alma is our product, so we say plainly where UWorld is the better pick.

How do UWorld and Alma compare at a glance?

Here is a quick factual comparison.

Feature UWorld (USMLE Step 1) Alma Study
Core purposeUSMLE style question banks and self assessments for Step 1, Step 2 CK and Step 3[2][3]AI study tutor that turns your own material into quizzes, flashcards, a schedule and Insights about what to study next
Question bankQuestion banks for Step 1, Step 2 CK and Step 3 with thousands of exam-style questions tested in an interface that mirrors the real exam[2][1]Questions generated from your own material in the style of the exam you choose (not real exam questions)
Self assessmentsUSMLE self-assessments that simulate the real exam interface, use the updated block format, and provide a predicted 3-digit USMLE score with a detailed score report[3]Does not include separate self-assessment exams or score prediction
Study plannerMedical planner that creates a daily schedule based on your availability to cover the subjects and systems you need to review, using QBank content[2]Study Schedule that builds a plan backward from your exam date using your topics and available hours
AnalyticsPerformance features that show peer-to-peer comparisons, including your overall percentile rank and average time per question compared to others[2]Insights dashboard with topic labels, forgetting forecasts, weekly retention and confidence calibration based on your own data
FlashcardsReadyDecks premade flashcards and SmartCards that are fully customizable and use spaced-repetition technology to increase retention and recall[2]Flashcard system with custom decks and spacing based on your review history and Learning Hub content
Mobile appMobile app that gives you reliable access to your full UWorld subscription, including QBank content, study tools and progress tracking, with automatic sync across devices[4]Web app accessed through your browser

What problem does each tool solve for a medical student?

UWorld and Alma solve different core problems. UWorld answers “What does the exam look like and can I answer those questions” with highly realistic vignettes and self assessments. Alma answers “Given my lectures, notes and current weaknesses, what should I study tonight, how should I review it, and am I actually learning or just re reading”. Using both sensibly means you stop guessing on either front.

Student choosing between a question bank door and a planning door.

What is UWorld built for?

UWorld is designed as an exam aligned question bank and self assessment suite for USMLE Step 1, Step 2 CK and Step 3.[2][3] Their USMLE prep is built on comprehensive question banks for each Step, with thousands of exam style questions delivered in an interface that mirrors the real exam.[2] For Step 1, every plan includes the full QBank, ReadyDecks flashcards, a Medical Library, UAsk AI Assistant, and personalized analytics.[1]

For self assessment, UWorld offers separate Step specific exams that simulate the real interface, question types, difficulty and block structure, and they give you an estimated 3 digit USMLE score, equated percent correct and peer comparisons.[3] These self assessments and QBank questions are written by practicing physicians and exam experts to deliver realistic testing.[3]

What is Alma built for?

Alma is not a question bank of real USMLE questions and does not simulate the official testing interface. Instead, it is a study tutor that turns your own PDFs, slides, text and decks into quizzes, flashcards, a schedule and an Insights dashboard. It is built to answer “what should I study next” and “is my studying actually working” using your own data, not generic percent correct in one QBank.

Where UWorld tells you how you perform on its questions, Alma groups your flashcards and quiz answers into topics, estimates mastery using its spaced repetition model, then feeds that into a Study Schedule that builds a plan backward from your exam date. It also serves nursing, NP and PA students, but everything in this article will focus on medical students.

How do UWorld and Alma decide what you should study next?

If you often log into UWorld, choose random unused blocks and hope coverage is “good enough”, this is the section that may change how you study. UWorld offers real-time performance analytics and peer-to-peer comparisons in its QBank, including what percentage of students selected each answer and your overall percentile rank.[2] Alma goes further on personalization: it builds topic level mastery from your flashcards and your own notebook-based quizzes, then uses that to pick the next cards, topics and calendar blocks.

Map of topics linked to a calendar.

How does UWorld guide what to study?

In UWorld you can create custom tests and quizzes from thousands of practice questions organized by subject and system.[2] In the Performance tab you can see your overall percentile rank and average time per question compared to others.[2]

For planning, UWorld offers a medical planner that creates a daily schedule based on your availability to cover the subjects and systems you need to review.[2][5] For many students this is enough in dedicated: run system based blocks in tutor and timed modes, then move to full self assessments to gauge readiness.

How does Alma Insights decide what is weak or strong?

Alma Insights takes a different approach. It does not start from a fixed exam outline. It starts from your own flashcards and Learning Hub quizzes and infers topics from those cards and questions. For each topic it combines:

  • Flashcard memory from its spaced repetition model (FSRS)
  • Quiz based understanding from Learning Hub quizzes
  • Recent “recall tests” from the last 30 days, meaning reviews of cards past the learning stage and graded quiz answers, plus optional Guess, Unsure or Sure taps

Every topic gets a mastery score and is labeled:

  • Needs work if mastery is under 50 percent
  • Getting there if mastery is 50 percent to under 75 percent
  • Solid if mastery is 75 percent or more
  • Too early to judge until there are at least 3 answers or reviews

A topic’s mastery is the lower of what Alma thinks you remember (flashcard memory) and how often you have actually been recalling it correctly recently. This protects you from a false sense of security on cards that are “on schedule” but still get failed often.

The Readiness tile is simply the weighted average of the mastery of all topics that have enough evidence. It only appears once at least three topics can be judged and it is explicitly not an exam score prediction.

What does “Do this now” and “About to forget” actually do?

Insights turns all those numbers into concrete actions in three important ways for medical students.

  • “Do this now” chooses a topic because your memory is fading, you have been missing questions on it, or it is one of your weaker topics. It tells you how many flashcards are due for that topic and roughly how long it will take, then lets you jump straight into reviewing those cards or asking the AI assistant about that topic.
  • “About to forget” lists topics whose flashcard memory is predicted to drop in the next week if you do not review. You can sort this list by expected weekly drop, or by lowest forecast memory on exam day once you have set a date.
  • “Is your studying working” charts, week by week, your flashcard retention and quiz accuracy alongside how many study events you had. Gaps show weeks with too little data. This makes plateaus and slumps visible.

On top of that, “Does feeling sure match being right” compares how often you were correct when you tapped Guess, Unsure or Sure. When there is enough data it flags patterns of overconfidence or underconfidence based only on your own answers.

Finally, Alma shows “Alma noticed” patterns once repeated wrong answers suggest the same confusion, for example consistently mixing up similar drugs or misreading a certain kind of stem. That is a prompt to adjust how you read or structure that material.

Where should a med student care most about this?

For Step 1 and Step 2 CK, the biggest waste of time we see is re reading notes or doing scattershot questions that feel busy but do not target genuine weaknesses. If you already have detailed percent-correct and subject-level feedback from a QBank, Insights is a layer on top that uses your own lecture-based questions and your flashcards to tell you which topics are slipping and how to spend tonight’s hour.

If you want more detail on how Insights works, we cover the mechanics in a separate guide: How Alma Insights actually finds your weak topics.

How do the study planners differ in practice?

UWorld and Alma both offer planners, but they play very different roles. UWorld’s study planner is an add on to its QBank. Alma’s Study Schedule is the core of how the app tells you what to do each day, using your own weak topics and hours. For busy clerkship students especially, this difference matters.

Exam calendar filled with study blocks.

What does UWorld’s planner do?

UWorld’s medical planner creates a daily schedule based on your availability to cover the subjects and systems you need to review, tied to its QBank.[2] From their copy the intent is clear: allocate questions in a structured way up to your exam, driven by QBank content.

How does Alma Study Schedule work for multiple exams and rotations?

Alma Study Schedule is a standalone calendar that can manage multiple exams and rotations at once. You tell it:

  • Your main exam name and date (for example Step 1 or Step 2 CK)
  • Other exams or shelf rotations and their dates
  • Your weekly study hours for each day and your days off
  • How many “catch up” days you want before each exam

What happens when life happens and you fall behind?

The reality of wards, calls and family is that your beautifully generated plan will break. Alma addresses this with two design choices:

For a deeper walkthrough, we have a full article on the planner here: How Alma’s Study Schedule actually builds your day by day plan.

How do question practice and explanations compare?

How strong is UWorld’s practice and rationale layer?

UWorld’s USMLE prep is built around question banks for Step 1, Step 2 CK and Step 3 that contain thousands of exam style questions, written by experienced doctors and delivered in an interface that mirrors the real exam.[2] Their answer explanations are heavily visual, with illustrations, diagrams, tables and educational objectives that make the tested concept explicit.[3]

On the Step 1 page they state that every plan includes the full Step 1 QBank, ReadyDecks flashcards, Medical Library and UAsk AI Study Assistant, along with personalized analytics.[1] They also describe survey data indicating that completing the UWorld Step 1 QBank at least once was positively related to Step 1 scores, although the original survey is not reproduced in our sources.[5]

UWorld also provides separate self assessment exams for each Step, with eight blocks of twenty unique multiple choice questions per test, simulating the updated USMLE format.[3] After a self assessment you receive an estimated three digit score, an equated percent correct and peer comparison breakdowns.[3]

What are Alma’s quizzes actually for?

  • Multiple choice quizzes written in the style of the exam you chose at sign up, with every option explained and a summary report of strengths and gaps
  • Mind maps that show how concepts connect

Within the Quiz tab you can choose 8, 12 or 15 questions, focus on particular sections, and later filter by questions you got wrong, right or have not seen yet. Each quiz ends with a report listing which areas were strong and which to revisit, and a per question breakdown of chosen answers and explanations.

Because these questions are written from your own lecture slides or reading, they are most valuable for aligning what you are told in class with how it might be tested, and for generating flashcards on the concepts you keep missing. They are not real or past exam questions and they do not predict a score.

How can you safely move missed UWorld questions into Alma?

For many students, the sweet spot is to use UWorld for exam quality questions and Alma for the downstream learning pipeline. A common workflow:

  1. Do a UWorld block in timed or tutor mode.
  2. For each missed question, summarize the key takeaway in your own words in a note, then use Alma flashcards to turn that into a card. For conceptual misses, consider an image occlusion card to hide labels on an anatomy diagram or pathway.
  3. Tag the card with the system and topic, for example “CV, valvular disease”.
  4. If several misses cluster around one lecture topic, paste parts of that lecture into Learning Hub and generate a quiz specifically on it to stress test your understanding.

This way, the heavy lifting of question quality and exam representation stays with UWorld, while Alma ensures that each UWorld miss becomes a spaced, trackable piece of memory work instead of a one off regret in your review history.

How do the flashcard systems and spacing models differ?

Both UWorld and Alma offer spaced repetition flashcards. UWorld provides ReadyDecks premade flashcards and customizable SmartCards that “feature spaced-repetition technology to increase your retention and recall.”[2] Alma’s flashcards are a core system that underpins Insights and the Study Schedule.

What does UWorld offer for flashcards?

With a USMLE QBank subscription, UWorld gives you access to ReadyDecks, premade flashcards that cover highly tested content for Step 1 and Step 2 CK, and SmartCards, which are customizable flashcards you can create from QBank content.[2][1] Both use spaced repetition to support retention over time.[2]

These cards are deeply integrated with the question bank, so they are especially useful if your study world lives mostly inside UWorld.

How does Alma use FSRS and why does that matter?

Alma’s flashcard system uses the FSRS (Free Spaced Repetition Scheduler), a modern algorithm that models memory using difficulty, stability and retrievability.[6][7][8] FSRS treats stability as how firmly a card is stored and retrievability as how easily you can recall it at a given time.[6] As time passes, retrievability drops following a specific forgetting curve, and the scheduler chooses the next review interval to hit a target probability of recall.[7]

In practice you see this as the familiar Again, Hard, Good and Easy buttons, each showing the next interval, and Alma adjusting review intervals so that more difficult cards are reviewed more often. The algorithm considers how complex the material is and how stable it already is, with stability increases becoming smaller as items become very well learned.[6]

Alma lets you import cards from your own decks, from the Learning Hub material, and from notes or PDFs. Image occlusion cards can be built for anatomy, histology and radiology images so that you hide labels and reveal them on flip. Everything you review feeds back into Insights and the Study Schedule.

If you want more detail on FSRS in Alma, we have a separate deep dive here: How Alma uses FSRS to schedule your flashcards.

What about Notebook, Practice Patients and the AI assistant layer?

Where UWorld distinguishes itself on QBank quality and self assessments,[2][1][3] Alma differentiates on breadth: Notebook for structured notes and mind maps, Practice Patients for clinical reasoning drills, and an AI assistant that answers from your own material and cites research for content questions.

How can Notebook help a med student stay organized?

Alma’s Notebook is centered on your own notes. You can build:

  • Mind maps across notes to visualize how concepts connect
  • Data tables and simple diagrams to straighten out comparisons like drugs or bugs
  • A knowledge graph like view of related topics
  • Flashcards from selected lines of text, which then enter the FSRS system
  • AI drawn images, either cartoon mnemonic style or textbook style, from any text description

Because Notebook is indexed, Alma’s AI assistant can answer questions by searching those notes first, grounding its answers in what you already decided to care about, rather than hallucinating generalities.

What do Practice Patients add that a question bank does not?

Practice Patients is Alma’s OSCE style feature for medical students, nurses, NPs and PAs. Stations are structured in five stages: history, physical examination, investigations, diagnosis and plan. At the end of a station you see marking that tells you which key steps you hit and which you missed. Some stations are available free each week.

These are not a replacement for UWorld style vignettes. Instead, they are a way to apply the same knowledge in an open ended, clinical reasoning format that more closely resembles ward thinking and OSCEs. That can be particularly helpful in the application and performance phases before Step 2 CK or shelf exams.

How trustworthy is Alma’s AI assistant?

Alma’s AI assistant does several things:

  • For study questions about your own material, it searches your Notebook, Learning Hub uploads and decks and bases answers on that content.
  • For medical content questions, it checks PubMed and cites sources so you can verify.
  • It never replaces guidelines or clinical decision support and is not for patient care.

This is important for safety: when you ask why an answer is wrong in a Learning Hub quiz, the “Ask Alma” button opens a chat seeded with the question and the explanation so far, rather than starting from scratch. You should still verify any clinical claim against a primary source when it matters.

If you want to think more deeply about when AI is appropriate in med school, see our guide: How to study from a textbook PDF with AI.

What does the learning science say about all of this?

Both UWorld and Alma lean on research backed principles: retrieval practice, spacing and metacognitive calibration. It is worth separating what has been shown generally from what has not been specifically tested for either product.

What the evidence does and does not show

Several lines of memory research are relevant:

  • Roediger and Karpicke showed that taking tests on material (retrieval practice) improves long term retention more than re studying the same material, especially at delays of two days to one week.[9]
  • Karpicke and Roediger later showed that repeated retrieval of already recalled items enhanced retention by more than 100 percent compared with dropping those items from further testing.[10]
  • The FSRS project formalizes spaced repetition in terms of stability and retrievability and provides algorithms for choosing optimal review intervals to maintain a desired probability of recall over time.[6][7][8]

These studies and models support the general strategy both tools rely on: test yourself often, space reviews instead of cramming, and focus especially on material you are about to forget.

However, there are important limits:

  • There is no published randomized or observational outcome study of Alma showing an effect on USMLE scores or pass rates. We cannot and do not claim Alma raises scores.
  • UWorld cites survey results suggesting associations between QBank completion and Step scores, but the primary data are not included in the sources here.[5] Associations are not proof of causation and may be confounded.
  • Alma’s Insights explicitly does not predict exam scores or pass chances. Its Readiness metric is an average of topic mastery from your own cards and quizzes, and it only sees what you study inside Alma.

Using either tool still requires judgment: tie your work to your school’s curriculum and official exam content outlines, monitor your well being, and do not treat any app as a guarantor of outcomes.

Who should pick UWorld only, Alma only, or both?

Most US students will use UWorld at some point because of its central role in Step prep. Alma is then a question of “also” rather than “instead of”. Here is a practical breakdown.

When is UWorld alone the right choice?

Consider UWorld only if:

  • You are in a short dedicated Step 1 or Step 2 CK window with limited budget and need exam style questions and self assessments above all else.
  • Your school already gives you an effective schedule and your own note and flashcard system that you trust, so you do not need an additional planner or Insights dashboard.
  • You are retaking a Step and want to focus almost entirely on QBank performance metrics and official NBME practice, not on building a long term topic map.

In these scenarios, put your money into UWorld, work through the QBank in a structured way, and use the self assessments and NBME forms to make timing decisions.

When might Alma alone be enough?

Alma alone may be reasonable if:

  • You are an early M1 or M2 still a year or more from dedicated, focusing on course exams and building a solid Anki or card base while you save for a QBank later.
  • You attend a school that provides an institutional QBank or comprehensive school license to another resource, and you want a personal tutor layer that is not tied to any one bank.
  • You are primarily struggling with planning, procrastination or fragmented notes rather than with question access.

Even then, when USMLE time comes, you will likely still want a major QBank like UWorld for genuine exam level practice.

When is using both UWorld and Alma worth it?

For many students, especially in the US, the answer is “both” for at least part of med school. Combining them is most compelling if:

  • You are in M2, starting to integrate systems across classes, and you want to keep your lecture notes, flashcards and early board style practice in one adaptive system.
  • You are in clerkships, juggling shelf exams and Step 2 CK, and need a planner that respects multiple exams and your work schedule while still letting you hammer UWorld for practice.
  • You know you tend to forget what you do outside of QBank apps unless it enters a spaced system, and you want Insights to tell you when key topics are sliding.

In this combined scenario, think of Alma as your daily home base: it tells you what to do today across cards, lecture based quizzes and weak topics. UWorld then fills the “performance” slots in that plan as your main source of exam style questions.

How to use UWorld and Alma together step by step

If you decide to use both, here is a concrete workflow you can adapt. Assume you are an M2 three months from Step 1, with classes in the morning and studying in the afternoon and evening.

  1. Set up Alma’s Study Schedule. Enter your Step 1 date, your average daily available hours by weekday, and any days off. Add current rotations or school blocks if relevant.
  2. Upload your most important lecture series into Learning Hub. For example, your school’s immunology and pharmacology notes. Generate mind maps and a first quiz for each to expose big holes early.
  3. Build core flashcard decks. Turn your best notes and Learning Hub explanations into flashcards. Use image occlusion for anatomy and path images.
  4. Start UWorld with system based blocks. Early on, choose tutor mode for a single organ system that matches your current lectures, for example renal.
  5. After each UWorld session, create Alma cards from misses. For every wrong or guessed right question, write a one sentence takeaway and convert it into an Alma card, tagging by system and topic.
  6. Let Alma Insights find the patterns. After a week or two of this loop, check Insights. Look at which topics are Needs work, what “About to forget” lists, and whether you are overconfident or underconfident.
  7. Adjust the Study Schedule. Hit “Catch me up” when wards or life knock you off plan. Pay attention to the Application and Performance phases, where Alma will start scheduling more mixed practice and review for your weakest topics.
  8. Use UWorld self assessments strategically. In the last month, place UWorld self assessments into the Performance phase blocks as your main score based checkpoints, and update your plan or even exam date based on those results and NBME guidance, not on Alma’s Readiness alone.
  9. Keep Notebook and Practice Patients for integration. When Insights or your own sense tells you that a topic is perennially weak, build a focused Notebook page and maybe a Practice Patient station list for it to practice application side, not just recall.

This setup respects what each tool is good at: UWorld for question realism and score tracking, Alma for structure, spacing and day to day decisions.

How should you judge whether Alma is worth adding?

Before you pay for any study tool, it is worth asking a few hard questions: what problem is it solving for you, what evidence supports its design choices, and how does it fit with what your school already gives you. That is as true for Alma as it is for UWorld.

If you want help thinking through those questions, we have a separate guide on how to evaluate AI study tools honestly: How to judge an AI study tool before you pay in med school.

Frequently asked questions

Does Alma replace UWorld for Step 1 or Step 2 CK?

No. UWorld’s USMLE question banks contain thousands of exam style questions and are built and updated by practicing physicians, with an interface that mirrors the real exam and self assessments that provide estimated scores and percentile rankings.[2][3] Alma does not have real exam questions or self assessments and does not predict scores. For most students Alma is a complement to, not a replacement for, UWorld.

Can I use Alma if my school gives me a different QBank instead of UWorld?

Yes. Alma is independent of any particular QBank. You can still turn missed questions from other banks into Alma flashcards, upload your lectures to Learning Hub for quizzes, and use Study Schedule and Insights to structure your time. You just will not be able to import questions directly; you summarize the takeaways yourself.

Is UWorld’s planner enough, or do I still need Alma’s Study Schedule?

For some students, UWorld’s medical planner, which creates a daily schedule to cover subjects and systems in the QBank based on your availability,[2][5] is sufficient. Alma’s Study Schedule adds advantages if you have multiple exams at once, want a three phase plan built around your own weak topics and flashcards, or know you will fall behind and need a “Catch me up” button that automatically replans. The right choice depends on how complex your exam calendar and life obligations are.


Alma is our product, so throughout this article we have been explicit about where UWorld is the stronger choice and where Alma adds value as a tutor and planner. Our goal is that you can make a clear, honest decision for your own situation.

References

  1. USMLE Step 1 Question Bank - UWorld Medical
  2. USMLE Study Material - UWorld Product Tour
  3. UWorld USMLE Self-Assessments - Determine Your Readiness
  4. USMLE App for Apple & Android Devices - UWorld Medical
  5. Best Resources for the USMLE - Steps 1, 2, & 3
  6. free-spaced-repetition-scheduler/README.md at main · open-spaced-repetition/free-spaced-repetition-scheduler · GitHub
  7. The Algorithm · open-spaced-repetition/fsrs4anki Wiki · GitHub
  8. fsrs - Rust
  9. Test-enhanced learning: taking memory tests improves long-term retention - PubMed
  10. doi:10.1016/j.jml.2006.09.004