How Many Anki Cards per Day for Medical School?

Disclosure: I build AnkiQuiz, the card-based quiz tool discussed below. AnkiQuiz is not affiliated with Anki, AnKing, AnkiHub, UWorld, AMBOSS, or the NBME. It does not replace a validated question bank, and AI-generated questions can contain errors.

Medical students often ask for one number: should you do 20, 50, or 100 new Anki cards a day? That question sounds precise, but the answer depends on the workload those cards create after today.

A new card is not one task. It can return in learning steps, become a review tomorrow, and remain in your schedule for months. The sustainable number is not the largest batch you can finish tonight. It is the number whose future reviews still fit beside lectures, clinical duties, question-bank work, sleep, and recovery .

This guide shows how to find that number using a time budget, a short baseline, Anki's workload tools, and a weekly quality check. It does not prescribe one universal card target. If you need the full study loop first, start with the daily Anki workflow for medical school .

The correct daily number can be zero during a backlog, a difficult rotation, or an exam-heavy week. Zero new cards does not mean zero learning; it means you are not creating an additional scheduled obligation that day.

New cards and reviews are not the same number

New cards are an input

Anki's New cards/day setting limits how many unseen cards can be introduced. It is an upper limit, not a goal and not a debt. The official Anki deck-options guide notes that studying fewer new cards does not cause the unused allowance to accumulate for the next day.

You do not need to “hit” the limit. Introduce only cards for material you have learned and intend to maintain.

Due reviews are the downstream workload

Reviews are cards the scheduler has brought back based on prior answers. Today's review count reflects earlier new-card intake, lapses, card difficulty, scheduling settings, missed days, and deck maturity.

Do not plan reviews by multiplying today's new cards by a fixed internet ratio. Two students introducing the same number can create different workloads because their cards, answer histories, desired retention, and learning steps differ.

Cards studied is not cards retained

A session total can mix new cards, learning repetitions, relearning steps, and mature reviews. “I did 500 cards” does not tell you how many distinct facts were learned or whether they can be retrieved under a different prompt. Use the count as a workload observation, not as a measure of understanding or exam readiness.

Start with time, not somebody else's card count

1. Protect the non-Anki work

Put these commitments into the day first:

• required classes, labs, and clinical duties;
• learning new material from your curriculum and trusted resources;
• legitimate question-bank blocks and detailed answer review;
• meals, exercise, sleep, and realistic transition time;
• school-specific assignments and assessments.

Anki can preserve knowledge represented on cards. It cannot replace first learning, clinical reasoning practice, or recovery. If the card target only works by deleting one of those activities, the target is already too high.

2. Give Anki a repeatable daily block

Choose a block you can maintain on an ordinary weekday, not only on a free Sunday. A rotation schedule may require a smaller weekday block and a separate weekend audit. A preclinical course may allow a longer, more predictable session. The exact duration is personal; what matters is that the block has a stopping point.

3. Measure seven normal days

For one week, record:

• minutes spent on due reviews;
• reviews completed and new cards introduced;
• Again responses and recurring leeches;
• unfinished due cards;
• question-bank or sleep time displaced.

Do not use your fastest day as the baseline. Use a typical pace that includes difficult cards and normal interruptions.

A capacity method that does not invent a magic number

Suppose you can reliably spend 60 minutes on Anki. During a seven-day baseline, due reviews take about 42 minutes on a normal day. That leaves roughly 18 minutes for learning new cards. Start with a deliberately conservative batch that fits those 18 minutes, keep it stable, and observe the next one to two weeks.

Do not convert those minutes using someone else's “seconds per card.” A short mature recognition card and a new management-algorithm card are not equivalent. Measure your own pace with your own material.

The delayed-workload test

After one to two weeks, ask:

• Is the due-review block getting longer?
• Are learning cards spilling into the rest of the day?
• Is the Again rate rising because cards were introduced too early?
• Are question blocks being shortened?
• Are you grading failed recall as Hard to keep moving?
• Is the queue still manageable on your busiest day?

If the answers are worsening, lower the new-card limit. Waiting until a backlog forms makes the adjustment more expensive.

Use Anki's workload tools

Check the forecast, but know its limits

Anki's statistics can show future due reviews. The forecast helps reveal whether a large unsuspension batch or recent intake is creating a wave. It is still an estimate: future failures and learning repetitions can add work, and overdue cards affect what the graph represents. Use it to compare plans, not to promise an exact future count.

Use the FSRS simulator

If you use FSRS, the Anki simulator can estimate reviews per day or study minutes using your cards' memory states, FSRS parameters, desired retention, new-card setting, and review limit.

Run it before a large change. Compare lower and higher new-card intakes over the period that matters for your course or rotation. A number that looks easy today may produce an unacceptable workload two weeks later.

Treat desired retention as a workload choice

The Anki manual explains that higher FSRS desired retention produces shorter intervals and more reviews. Its default 90% is presented as a balance, and workload rises quickly above it. Do not raise retention, unsuspend hundreds of cards, and increase new cards at the same time. Change one variable and observe the result.

See the full FSRS settings guide for medical school before changing learning steps, desired retention, optimization, or rescheduling.

Should maximum reviews per day be 9999?

Advice to set Maximum reviews/day very high is intended to prevent a low cap from silently hiding due cards. It does not mean every medical student must finish an unlimited queue regardless of cost.

A review cap changes which cards Anki shows; it does not remove the underlying forgetting or repair an oversized active deck. If the cap repeatedly binds:

1. pause or reduce new cards;
2. inspect whether low-value subdecks are feeding the parent deck;
3. repair or suspend leeches and duplicates;
4. protect newly due, high-value material;
5. make a deliberate plan for older overdue cards.

If you already have a backlog, use the medical-school Anki backlog recovery plan instead of changing a visible limit and assuming the workload disappeared.

Adjust the number by training stage

Preclinical courses

New cards often follow the material taught that day. Preview the relevant AnKing tags or your own cards, learn the topic first, and activate the smallest set that tests useful relationships. Do not force the same number every day: a fixed quota can encourage low-value cards just to reach the target.

Step 1 dedicated

Question blocks, answer review, and official assessments become more important. Broad new-card intake often needs to fall sharply or pause. Add or reactivate cards for documented, reusable knowledge gaps rather than trying to mature an untouched deck. The Step 1 dedicated Anki workflow explains how to bound reviews without treating a generated quiz as a readiness measure.

Clerkships and shelf exams

Clinical days are variable, so set intake against the busiest repeatable weekday. Use current shelf or rotation tags as candidate sets, preview them, and activate material as you learn it. Re-triage at each rotation change so old specialties do not leave their full workload behind. See the clerkship and shelf-exam Anki workflow .

After a missed day or during a backlog

Set new cards to zero temporarily. Clear due work in bounded sessions and restart at a smaller intake only when the queue fits again. Do not add the “missed” new-card allowance to tomorrow; Anki does not require that.

Card quality changes daily capacity

Overloaded cards

A card asking for a diagnosis, mechanism, treatment, contraindications, and adverse effects at once is difficult to grade. If you recall four of five parts, neither Again nor Good accurately describes what happened. Split the prompt into the few relationships worth scheduling and keep supporting explanation in a non-tested field.

Vague prompts

“What is the treatment?” has no unique answer without a condition, stage, severity, or patient context. Add the minimum context required for one defensible target.

Sibling cueing

Several clozes from the same sentence can become easy because the visible wording identifies the hidden phrase. Occasionally ask the relationship in the opposite direction or under different wording. If the changed prompt fails, repair the source card rather than merely increasing daily volume.

Leeches and interference

The Anki manual defines leeches as cards you repeatedly forget and notes that they consume disproportionate review time. Search tag:leech regularly. Relearn the concept, clarify the prompt, separate similar items, add a useful mnemonic, suspend low-value material, or delete a card that is not worth maintaining.

Use AnkiQuiz as a quality check, not another daily quota

Normal Anki reviews ask whether you can retrieve the target when this card appears on schedule. They do not always show whether the familiar cloze, layout, or neighboring text has become part of the cue.

AnkiQuiz can turn selected card text into single-choice, multiple-choice, fill-in-the-blank, or true/false questions. Its useful role here is a small weekly changed-prompt audit:

1. Select 15–30 mature cards from one current topic.
2. Exclude tags, source IDs, links, protected question-bank content, and sensitive information.
3. Generate fewer questions than source cards.
4. Ask for reworded prompts that stay within the supplied facts.
5. Answer without opening the original cards.
6. Verify answers against the cards and a trusted current medical source.
7. Classify misses as cue dependence, weak understanding, or interference.
8. Repair the cards or relearn the topic before increasing intake.

The card-to-quiz guide explains installation, field selection, and generation.

AnkiQuiz does not mark cards reviewed, update FSRS, change due dates, or determine your safe daily intake. Generated questions can be ambiguous or wrong, and context-poor cards cannot reliably become validated USMLE- or shelf-style vignettes.

Warning signs that your daily number is too high

Lower or pause new cards when:

• due reviews exceed their time block for several days;
• question-bank work or answer review is repeatedly skipped;
• you introduce cards before learning the topic;
• you press Hard on failed recall to avoid relearning;
• leeches and ambiguous cards are accumulating;
• most reviews are rapid pattern recognition;
• sleep, exercise, or clinical preparation is being sacrificed;
• a single missed day creates an unmanageable queue.

The useful response is usually to reduce input and improve scope or card quality, not to click faster. Use the medical-school card quality checklist to repair slow, ambiguous, or cue-dependent prompts.

Questions medical students commonly ask

How many new Anki cards should I do per day?

There is no universal evidence-based number. Set a daily Anki time budget, measure how much of it due reviews use, and introduce a conservative number of new cards in the remaining time. Reassess after one to two weeks because the cost is delayed.

How many Anki reviews per day is too many?

The count is too high when honest reviews no longer fit beside required learning, question practice, clinical duties, and sleep. Review density and card quality differ, so another student's count is not a valid threshold.

Should I finish all reviews before new cards?

Due reviews usually deserve priority because new cards add future work. During a large backlog, protect newly due high-value reviews and use a bounded catch-up plan rather than allowing old cards to consume all course or question-bank time.

Is 100 new Anki cards a day too many for medical school?

It may be sustainable briefly for some students and material, but the relevant test is the downstream workload. Simulate it, measure it for at least one to two weeks, and reduce it if reviews, failures, or displaced work rise.

Should I set new cards to zero during dedicated?

Zero can be appropriate when broad intake would displace question blocks or when you already have enough active material. Selective new or unsuspended cards may still help a documented knowledge gap. Follow exam-directed needs, not a fixed rule.

Does AnkiQuiz reduce my Anki review count?

No. It is a separate quiz-generation workflow and does not alter Anki scheduling. Use it occasionally to check whether mature card knowledge survives changed wording, then make necessary repairs in the source deck.

Choose a number that survives next week

The best daily Anki target is not the number that makes today feel productive. It is the number that still leaves a manageable queue after busy days, difficult topics, question blocks, and ordinary life.

Protect the work Anki cannot replace. Measure your own review cost. Introduce new cards conservatively, inspect the delayed workload, and audit whether familiar prompts are hiding weak retrieval. When the system stops fitting, reduce the input before motivation and sleep become the limit.