If you've talked to a single MCAT studier in the last decade, you've heard the word Anki. Maybe you've already downloaded a 25,000-card premade deck and watched the daily review queue grow into something genuinely terrifying.
This is a piece about whether that's the right approach, what spaced repetition actually does for MCAT content, and what an alternative looks like that doesn't require you to become a part-time deck manager.
I want to be careful here. Anki works. Spaced repetition works. The question is whether the specific way most pre-meds use it — premade megadecks, custom card creation, manual scheduling — is the best use of the limited months you have before test day.
What spaced repetition is doing for you
Spaced repetition exploits a finding from memory research that's about a hundred years old at this point: information you retrieve right before you would have forgotten it sticks dramatically better than information you re-read while it's still fresh.
The practical implication: a flashcard you got right yesterday should not appear today. It should appear in three days. If you get it right again, in a week. Then two weeks. Then a month. Each successful retrieval pushes the next exposure further out.
This works because retrieving a memory at the edge of forgetting is the strongest possible signal to your brain that this is information worth keeping. Re-reading the textbook chapter for the third time isn't sending that signal at all.
For MCAT content, this is enormous. The test is a vocabulary test wearing a science-test costume. You need to recognize roughly 2,000 to 4,000 specific concepts, terms, pathways, equations, and relationships. The half-life of any of those, if you study them once, is about a week. If you space them, the half-life starts pushing past test day.
So the question isn't whether spaced repetition belongs in MCAT prep. It does. The question is how.
The Anki problem
Anki is a fine piece of software. It's open source, it's free, and the underlying algorithm (a variant of SM-2) is solid. The issue isn't the tool. The issue is the workflow most pre-meds end up in.
The premade deck trap. You download a 25,000-card deck someone else made. You commit to "doing 200 cards a day." Within three weeks the daily queue has somehow grown to 600 cards because of all the relearning, you're spending two hours a day on cards alone, and you haven't read a chapter of biochem in a week. You're not actually learning the material — you're recognizing the cards.
The card-quality problem. Premade decks vary wildly in card quality. Some cards have ambiguous answers. Some test the wrong thing. Some assume context you don't have because you haven't read that chapter yet. You spend mental energy parsing whether the card is bad or you're wrong.
The deck-creation tax. The alternative — make your own cards — is even more expensive. You're spending an hour a day writing cards instead of studying. The cards are good, because you understood the source material, but you're trading study time for card creation time.
The all-or-nothing burnout. Skip three days because you had a hard week, and now your queue has 1,800 cards waiting. The psychological weight of that is real. A lot of people just stop opening Anki at this point and pretend it didn't happen.
This is the failure mode I want to talk about, because it's specifically what Fennie was built to avoid.
What MCAT content actually demands
Let's be specific about what you're trying to memorize. The MCAT covers four sections, each with predictable content categories.
Bio/Biochem (60% bio, 40% biochem). Amino acids — structures, properties, classifications. Enzyme kinetics. Glycolysis, citric acid cycle, electron transport chain — every step, every enzyme, every cofactor. Lipid metabolism. DNA replication and translation. Cell biology. Genetics.
Chem/Phys. Stoichiometry, thermodynamics, kinetics. Acid-base chemistry. Redox. Some organic chemistry but less than people fear. Plus mechanics, fluids, electricity, optics — physics taught at a level that mostly tests application of formulas.
Psych/Soc. Often called "the easy section" by people who haven't studied for it. Roughly 600 specific terms — defense mechanisms, social theories, neurotransmitters and their associated systems, learning theories, cognitive biases. The volume is the issue.
CARS. No content to memorize. A skills section.
So three of the four sections are fundamentally vocabulary-and-relationships sections. Spaced repetition is suited for them. CARS isn't a flashcard problem.
For each of the three content sections, you're looking at maybe 1,000 to 1,500 atomic facts you need to recognize on demand under timed pressure. That's a real spaced repetition workload — but a manageable one if you don't drown yourself in 25,000 cards of varying quality.
The note-first alternative
Here's a different model that I think most pre-meds would benefit from.
You read or watch the source material. You take notes — markdown is fine, your own words are required. You generate flashcards from the notes. The system schedules them.
The key move is: the cards come from the notes, not from a stranger's deck. This solves three problems at once.
First, you don't write the cards manually. The note already contains the content. The flashcard generator pulls definitions, relationships, and key facts out of the note. You skim the cards and edit any that are off — usually a 30 second pass.
Second, the cards are cards you made (in effect), so they match how you understand the material. The deck creator isn't a stranger from 2018 who wrote unclear cards.
Third, because the cards are tied to the notes, you can see what you're studying. A card that's been failing for three weeks pulls you back to the source note for re-reading. The card and the source live next to each other.
This is the Fennie model. You take notes — even rough lecture notes are fine. From any note, you can generate a flashcard deck in two clicks. The cards go into Memory's spaced schedule. They show up in your daily plan when they're due.
There is no card creation tax, because the cards come from your notes. There is no premade deck quality issue, because the cards reflect your understanding. And there is no daily queue panic, because the daily plan caps the number of due cards based on the rest of your day.
A realistic MCAT week
Here's what a week looks like for a pre-med running this kind of approach, three months out from test day.
Monday-Friday during a content block. Read the chapter (60 min). Take a clean note in your own words (20 min). Generate a flashcard deck from the note (2 min, mostly reviewing the cards). The note plus deck takes about 90 minutes for the chapter. Then 20 to 30 minutes of due cards from prior weeks' decks. Total: roughly 2 hours daily.
Saturday. Half-length practice section. The structure of the test starts mattering — you can't just know the content, you have to apply it under time pressure. Plus 30 minutes of due flashcards.
Sunday. Light. Catch up on flashcards, review the practice section's mistakes. 90 minutes total.
That's about 12 to 14 hours a week of structured prep, which is roughly what most successful MCAT studiers report. The key is that the spaced repetition burden is integrated into the daily plan rather than being a separate, nightmarish, ever-growing queue.
Memory's role in this
Fennie's Memory system tracks a few things about your flashcard performance that matter for the MCAT specifically.
Topic-level mastery. It's not just card-by-card. Memory rolls up performance to the topic level. If your amino acid cards are reliably failing, it knows amino acids are weak — not just one specific card. The daily plan can then surface a lesson on amino acids, not just more cards.
Forgetting curves per topic. Some topics you forget faster than others. Memory tracks this and adjusts spacing per topic. Cards on topics you forget fast come back sooner. Cards on topics that stick come back later. You don't tune anything — it just adjusts.
Cross-section connections. A glycolysis card and a metabolism essay question are both metabolism. Memory connects them. If you're failing the glycolysis cards, the system knows your metabolism essay performance is also at risk and surfaces both.
This is the thing that's hard to do in Anki without massive plugin work. Anki tracks individual card performance. It doesn't know the cards belong to topics, it doesn't know about your essays or quizzes, it doesn't know your test is in three months and the schedule should bias toward "everything reviewed at least twice before then."
What this doesn't replace
I want to be honest about limits.
A spaced repetition system makes the vocabulary half of MCAT prep efficient. It does not replace:
- Practice tests. You need full-length AAMC tests. They are the gold standard for both content and pacing.
- CARS practice. Reading hard passages under time. No flashcard helps.
- Application practice. Knowing the citric acid cycle is different from solving a passage problem about it.
- Office hours / tutors / study groups. Some content needs a human to explain it the third time.
The honest model: spaced repetition handles roughly 60% of your content workload very efficiently. The other 40% — application, integration, pacing, weak-area diagnosis — needs different tools. Fennie's plan layer integrates those (lessons, quizzes, calendar-aware spacing) but you're going to use AAMC's own materials too, and you should.
The dial-it-down version
If reading this made you anxious about your existing Anki workflow, take a breath. The fix is small.
Stop adding new cards from premade decks. Start writing notes for the chapters you're about to study. Generate cards from those notes. Let the existing Anki queue play out, but stop feeding it.
Or, frankly, switch tools. Fennie's flashcards live inside the daily plan and connect to your notes. You don't manage a deck; the deck is a byproduct of your notes. The schedule is set by Memory based on your performance and your test date.
The point is to use the science of spaced repetition without making yourself a part-time librarian.
Build your first MCAT note and watch the cards write themselves. Start with Fennie free — no card, no commitment.