How to Make Flashcards That Actually Stick: A Spaced Repetition Guide

Flashcards have a reputation problem. Plenty of students make hundreds of them, review them for a week, and then quietly abandon the deck when reviews pile up and nothing seems to stick. The method is not broken. Usually the cards are.

This guide covers the two halves of the problem: how to write cards your memory can hold on to, and how to schedule reviews so the habit survives a full semester. It works for any subject, any app, and any source material, from textbook chapters to recorded lectures.

Why flashcards work (when they work)

Three well-studied effects from memory research explain why flashcards beat rereading.

• The forgetting curve. Hermann Ebbinghaus showed in the 1880s that memory for new material drops off quickly at first and then more slowly. Without review, most of what you learn in a lecture fades within days.
• The spacing effect. Reviews spread out over time produce stronger long-term memory than the same number of reviews crammed together. Each well-timed review also slows down how fast you forget next time.
• The testing effect (retrieval practice). Pulling an answer out of memory strengthens it more than looking at it again. Research such as Roediger and Karpicke's 2006 study found that students who practiced recall remembered more a week later than students who spent the same time restudying, even though restudying felt more productive.

A flashcard combines retrieval with a schedule. The front forces recall; the app decides when you see it again. If either half is weak, the system leaks.

The minimum information principle

The single most important rule for writing cards comes from Piotr Wozniak, the creator of SuperMemo, in his widely shared list of rules for formulating knowledge: make each card as simple as possible.

Simple cards win for practical reasons:

• You either know the answer or you do not, so grading is honest.
• The scheduler can learn how hard that one idea is for you, instead of averaging across five ideas on one card.
• Failing a small card costs seconds. Failing a paragraph card is demoralizing, and demoralizing cards are how decks die.

Simple does not mean shallow. You can still learn complex topics; you just break them into many small, connected cards rather than a few giant ones. Understanding comes first, then the cards lock in the pieces.

Card types and when to use each

Most spaced repetition apps support four main card types. Here is what each one is good for, with a bad and a good example.

A few patterns are worth calling out:

• Lists are the enemy. "Name the five stages of grief" is a hard card forever. Use one cloze per item, or a card per item with a cue ("Stage that follows denial?").
• Ask "why" and "how", not just "what". A card like "Why does increasing interest rates tend to reduce inflation?" builds understanding, as long as the expected answer stays short.
• Avoid yes/no questions. You can guess them, which makes grading meaningless.
• Add just enough context. "What is the capacity?" is ambiguous a month later. "What is the default capacity of a Java ArrayList?" is not.

Step-by-step: turning lecture notes into cards

Here is a workflow that works whether your source is a textbook chapter, a slide deck or a recorded lecture.

1. Understand before you memorize

Do not make cards during the first pass. Read, watch or attend first. If you cannot explain a concept in your own words, a card will only help you memorize words you do not understand. Make a note of anything confusing and resolve it before writing cards.

2. Extract the atomic ideas

Go through your notes and list every fact, definition, relationship and procedure worth remembering, one per line. This is the step most people skip. A list like this is much easier to turn into good cards than a page of prose.

3. Choose what not to learn

Not everything deserves a card. Ask: will I need to recall this without looking it up? Exam-tested definitions, core formulas and key relationships, yes. Trivia your professor mentioned once, probably not. Fewer, better cards beat a bloated deck.

4. Write one card per idea

Pick the card type that fits each line. Keep answers short, ideally a word, a number or a single phrase. If an answer needs a sentence, check whether it is really two cards.

5. Tag and source everything

Tag cards by course and chapter or lecture date. When a card stops making sense in three months, you want to know where it came from. Tags also let you filter study sessions before a specific exam.

6. Review the same day

The first review is the most important. Studying new cards the day you make them, while the material is fresh, sets up every later interval.

How spaced repetition scheduling works

Every time you review a card, you grade how well you remembered it. The app uses that grade to decide when to show it next: further out if it was easy, sooner if you struggled, and back to the start if you forgot it.

Anki, the most popular free option, ships with two scheduling algorithms.

Practical advice: if you are starting fresh in a current version of Anki, turn on FSRS in your deck options, keep the default desired retention, and run the parameter optimizer once you have a few weeks of reviews. If you already have a large SM-2 deck that works for you, there is no emergency to switch.

Whatever the algorithm, the grading buttons only work if you are honest. Pressing "Good" on a card you half remembered trains the scheduler to show it too rarely.

A daily routine you can sustain

Spaced repetition is a long game. The schedule assumes you show up most days. Here is a routine that holds up during a busy term.

1. Do due reviews first, every day. Reviews are the backlog that matters. New cards can wait; overdue reviews compound.
2. Cap new cards. Start with 10 to 20 new cards a day. Each new card generates future reviews, so your steady-state review count ends up several times your new-card rate.
3. Split sessions. Two 15-minute sessions (for example morning and evening) often feel easier than one 30-minute block, and phones make this easy.
4. Plan for exam crunch. Add cards for a course steadily across the term. Dumping 600 cards the week before an exam defeats the spacing effect.
5. Handle missed days gently. If you miss a few days, do not try to clear everything at once. Pause new cards and work down the backlog over several days.

A useful check: if your daily reviews regularly take longer than you are willing to spend, lower new cards per day or delete low-value cards. Do not stop reviewing.

Using AI to draft cards (and how to vet them)

AI tools are now good at turning notes, slides or lecture transcripts into a first draft of a deck. Used well, this saves a lot of typing. Used carelessly, it produces hundreds of trivia cards you never needed.

A prompt that tends to produce usable drafts:

You are helping me make spaced repetition flashcards from my study material.

Rules:
- One idea per card. Answers should be one word, number or short phrase.
- Skip trivia, anecdotes and examples unless they are needed to understand a concept.
- Prefer "why" and "how" questions for key concepts, with short answers.
- Use cloze deletions for facts that sit naturally in a sentence.
- Never make a card that asks for a list. Split lists into one card per item.
- Include enough context in each question that it makes sense on its own months later.
- Output as a table: Type | Front | Back | Source section.

Material:
[paste your notes or transcript here]

Then vet every card before you study it:

• Check facts against the source. AI can misstate numbers, dates and definitions. If a card is wrong, you will memorize an error.
• Delete aggressively. Expect to cut a meaningful share of a draft. Trivia and duplicates are common.
• Split compound cards. Any answer with "and" in it is a candidate.
• Rewrite in your words. Cards you have phrased yourself are easier to recall and easier to understand later.
• Keep it to material you understand. If you cannot explain why an answer is right, go back to the source first.

If your study material lives in recorded lectures or videos, getting a clean transcript first makes this step much faster. YouTubeTranscript.dev's Study tool can turn a lecture video into AI notes and flashcard drafts you can then edit and import.

Common mistakes that kill decks

• Making cards before understanding. You end up memorizing phrasing instead of meaning.
• Long answers. If you need more than a few seconds to say the answer, the card is too big.
• Lists and enumerations. Split them or use one cloze per item.
• Ambiguous questions. If two answers could be right, add context until only one is.
• Copying text verbatim. Textbook sentences make bad cards. Rephrase into a question.
• Too many new cards. The review avalanche arrives two to three weeks later, which is when most people quit.
• Skipping days, then binge-reviewing. Consistency matters more than session length.
• Never editing. Cards you fail over and over (often flagged as "leeches") are usually badly written. Rewrite, split or delete them.
• Grading generously. Honest grades are what make the schedule accurate.

A quick card-writing checklist

Before you add a card, run it through these questions:

1. Does it test exactly one idea?
2. Is the answer short and unambiguous?
3. Would it make sense if I saw it with no context in six months?
4. Do I understand why the answer is correct?
5. Will I actually need to recall this?

If you can answer yes to all five, the card is probably worth keeping. Write fewer, better cards, review every day, and let the schedule do the rest.