How to actually use flashcards (and when not to)

By Planr.ai · 8 September 2026 · 7 min read

Almost everyone who says flashcards did not work for them made the same three mistakes: cards that were too big, review that happened once, and a deck that was never used for anything except being made.

Making the deck is not studying

The most common flashcard failure is spending two hours building a beautiful deck the night before a test and calling that revision. Typing a definition onto a card moves it past your eyes, not into your memory. The work that moves it is pulling it back out, cold, when the answer is not in front of you.

A rough rule: if you spent more time making the deck than reviewing it, the deck was a procrastination project. Making 40 cards should take about 25 minutes. Reviewing them properly over the following week takes about 90 minutes in total, spread across short sessions. The ratio should be at least three to one in favour of reviewing.

What a good card looks like

A card should test one fact, and the answer should be short enough to say out loud in about five seconds. The moment a card has three bullet points on the back, you stop testing recall and start rereading a paragraph with extra steps — you glance at it, think that looks familiar, and flip.

Take a chunky card like: what caused the 1929 crash? The answer is a whole essay, so you never actually answer it. Split it: what was buying on margin? What proportion of a share price could an investor borrow in 1929? Name two warning signs that appeared before October 1929. Each is answerable in a breath, and each one either comes back or does not — no fuzzy middle where you tell yourself you knew it.

One fact per card, answerable in about five seconds.
Ask a question rather than showing a term — Define osmosis beats Osmosis.
Put the hard direction on the front: if you need to produce the Spanish word, the English goes on the front.
Use your own words on the back, not the textbook's. If you cannot phrase it yourself, you do not understand it yet, and the card will only teach you a sentence.
Add a two-word context tag when a term means different things in different units — Enzymes (digestion) is a different card from Enzymes (metabolism).

The spacing that actually matters

The research on spaced repetition is one of the sturdiest findings in learning science, and the practical version of it is simpler than the apps suggest. Review a card the day you make it, again two days later, again a week later, and again two weeks later. Four passes with widening gaps beats eight passes in one evening, and takes less total time.

If a test is close, compress the same shape rather than abandoning it: today, tomorrow, the day after, and the morning of. The gaps matter more than their exact length. What you are avoiding is the massed review where you see a card five times in ten minutes, get it right five times because it is still sitting in your short-term memory, and learn nothing.

One rule does most of the work: a card you got right goes to the back of the pile, a card you got wrong comes back in the same session and again tomorrow. That is the whole algorithm. Everything an app adds to it is refinement.

Say the answer before you flip

Out loud, or at least under your breath. The gap between thinking I know this and producing the sentence is where most exam marks are lost, and flipping fast hides that gap from you. Saying it forces a full answer and makes the near-misses obvious — you hear yourself get two of the three parts.

It is also the fastest honesty check available. If you cannot say it, you do not know it, whatever the feeling of familiarity is telling you.

Three subjects where cards are the wrong tool

Maths and physics problem-solving. Cards can hold the formulas, and they should, but the exam asks you to choose a method under pressure. That skill only comes from working problems. A deck of formulas plus twenty worked problems beats a deck of sixty cards.

Essay subjects at the argument level. Cards are good for dates, definitions, quotations and case details — the raw material. They are useless for the thing being marked, which is whether you can build a line of argument across four paragraphs. For that, write plans.

Anything you do not understand yet. Flashcards are a memory tool, not a comprehension tool. Making cards for a topic you cannot follow produces a deck of sentences you can recite and cannot use, which then fails the first question that phrases it differently. Understand it first, in any form — a worked example, a diagram, five minutes of a teacher explaining it — then make the card.

A realistic weekly rhythm

Make cards the same day you cover the material, while the lesson still makes sense, and cap it at ten or fifteen cards per lesson. Review yesterday's cards plus a random handful of older ones for ten minutes at the start of each study session. Do a full pass over a unit's deck the weekend before its test.

Ten minutes a day across a term is roughly ten hours, and it is spread so thin that you never feel it. The equivalent crammed into the week before finals is the same ten hours, except it collides with four other subjects doing the same thing.

Planr can build a deck from notes or a photo of a worksheet in about a minute, which removes the typing but not the judgement — you still have to split the big cards, and you still have to review them. The tool that makes the deck is not the part that learns it.