Spaced Repetition for Vocabulary: The Science of Remembering Words

You look up a word, understand the sentence, and feel a small spark of progress. Then a week later the same word appears and it’s a total stranger. That gap between meeting a word and owning it is where most vocabulary study quietly falls apart. Spaced repetition is the tool built specifically to close it — a way to schedule reviews so words move into long-term memory instead of leaking back out. Here’s how it works and how to use it well.

What is spaced repetition?

Spaced repetition is a study method that spreads your reviews of a word out over increasing intervals of time instead of cramming them together. You see a word today, again in a couple of days, then a week later, then weeks after that. Each successful recall pushes the next review further out, so memory strengthens with less total effort.

The core idea is almost boringly simple: memory is a use-it-or-lose-it system, and the timing of use matters as much as the amount. Reviewing a word ten times in one sitting does surprisingly little. Reviewing it ten times across several weeks — each session placed just as the memory starts to fade — does a great deal. Spaced repetition is the deliberate practice of choosing those moments.

Most people already do a crude version of this without naming it. You reread a note before an exam, glance at a name before meeting someone again, replay a song lyric until it sticks. Spaced repetition systems, or SRS, take that instinct and make it precise, using an algorithm to track hundreds or thousands of words and tell you exactly which ones need attention on any given day.

When people talk about spaced repetition for language learning, they usually mean one of two things: a dedicated flashcard app that schedules your reviews, or a built-in review feature inside a reading or study tool. Both rest on the same principle. The difference is mostly in how much friction sits between reading a word in the wild and getting it into your review queue.

Why do we forget words so fast?

We forget words fast because forgetting is the brain’s default, not a malfunction. New information decays quickly unless something signals that it’s worth keeping. A word you meet once, with no emotional weight and no repetition, is exactly the kind of thing your memory is designed to discard to make room for what you actually use.

The classic description of this pattern is the forgetting curve, associated with the 19th-century psychologist Hermann Ebbinghaus. He studied his own memory of nonsense syllables and described how retention drops sharply soon after learning and then continues to fade more gradually over time. The exact shape depends on the material and the person, so it’s best understood qualitatively: memory for something new tends to fall off quickly at first, then more slowly, unless you intervene.

Ebbinghaus also noticed the flip side, which is the part that matters for learning. Each time you successfully review something before it’s fully gone, the curve gets shallower — the memory decays more slowly than it did the last time. Forgetting doesn’t reset you to zero; it leaves a trace that’s faster to rebuild. Spaced repetition is essentially a strategy for exploiting that: catch the word on its way down, refresh it, and each refresh buys you a longer stretch of remembering.

Why isolated words fade fastest

Not all words forget at the same rate. A word tied to a vivid situation, a strong feeling, or a rich web of associations sticks far better than one memorized in isolation. This is why a word you looked up in the middle of a gripping story often survives longer than a word from a bare vocabulary list — the story gave your brain more hooks to hang it on.

Several factors speed up forgetting:

• No context. A word with no sentence around it is just a sound attached to a translation, easy to confuse and easy to lose.
• No emotional or personal relevance. Words that don’t connect to anything you care about have little reason to stay.
• Interference. Similar-looking or similar-sounding words blur together, especially when you learn a batch of them at once.
• Single exposure. One meeting is rarely enough for anything but the most memorable words.

Understanding why forgetting happens tells you what to fight against. You want more context, more meaningful connection, and above all more well-timed repetitions. Spaced repetition handles the timing; how you build your cards handles the rest.

How does spaced repetition help you remember?

Spaced repetition helps by scheduling each review for the moment your memory of a word is starting to weaken but hasn’t collapsed. Recalling something at that point of mild difficulty strengthens it far more than easy review. Over time the intervals expand — days, then weeks, then months — and the word settles into durable long-term memory.

Three mechanisms do the heavy lifting. Understanding them helps you use any SRS well, and forgive it when it feels frustrating.

Reviewing right before you forget

The magic timing is the edge of forgetting. If you review a word while it’s still fresh and effortless, you learn almost nothing — there was no struggle, so no signal that the memory matters. If you wait until it’s completely gone, you’re relearning from scratch. The sweet spot is in between, where recall takes a little effort but still succeeds. A good SRS tries to place each review in that window.

Expanding intervals

Because each successful recall makes a memory more durable, the next review can safely wait longer. So the intervals grow: maybe one day, then three, then a week, then two weeks, then a month, and onward. This expansion is what makes spaced repetition efficient at scale. A word you know well might not come up for months, freeing your daily reviews for the words that actually need work. You’re not reviewing everything constantly; you’re reviewing each thing exactly as often as it needs.

Active recall

Spaced repetition almost always pairs with active recall — forcing yourself to produce the answer before checking it, rather than just rereading. Trying to remember a word’s meaning, and briefly failing or straining, is itself a powerful learning event. This is why flashcards work and rereading a list doesn’t: the act of retrieval, not re-exposure, is what carves the memory deeper. The spacing decides when you retrieve; active recall is what makes each retrieval count.

It’s worth resisting the urge to peek at the answer too quickly. The value is in the attempt, even a failed one — the few seconds of genuine effort to dredge up a word do more for retention than immediately flipping the card. Give yourself a real beat to reach for it. If nothing comes, that’s useful information too: it tells the system this word still needs work, and it primes your brain to grab the answer more firmly when it appears. A slightly uncomfortable pause is the sound of learning happening.

Put together, these three ideas explain why spaced repetition can feel almost too effective. You’re not working harder — often you’re doing fewer total reviews than a crammer — but every review is placed where it does the most good. If you want a deeper look at how this fits into a reading-first approach, see our guide on learning by reading.

Spaced repetition vs. cramming — what’s the difference?

Cramming packs many repetitions into one short session; spaced repetition spreads fewer repetitions across many sessions over time. Cramming produces a quick, shallow memory that fades within days. Spacing produces slower-feeling gains that last for months or years. For anything you want to keep, spacing wins decisively on retention per unit of effort.

The frustrating truth about cramming is that it feels more effective than it is. Massed practice makes information fluent and easy in the moment, and that fluency tricks you into thinking you’ve learned it. Spacing feels harder — you keep half-forgetting things — and that difficulty is precisely the sign it’s working. Researchers sometimes call this a “desirable difficulty”: the struggle is the mechanism, not a bug.

Cramming isn’t useless — if you have a test tomorrow and will never need the material again, massing your practice is a rational choice. But vocabulary is the opposite of a one-off. You want these words available years from now, in real conversations and real reading. That’s exactly the case spaced repetition is built for.

How do you actually use spaced repetition for vocabulary?

In practice you build small cards — each one a word or phrase with its meaning — and review them daily in short sessions. The app shows you a prompt, you try to recall the answer, then you rate how hard it was. That rating tells the algorithm when to show the card next. You add a handful of new words each day and let the schedule do the rest.

Here’s the loop in more concrete terms.

Building the card

A card has a front (the prompt) and a back (the answer). For vocabulary, the most durable cards put the word inside a real sentence rather than stranded alone. A good card might show the sentence you found the word in, with the target word highlighted, and reveal the meaning on the back. This gives you context on every single review, which fights the isolation problem we covered earlier.

Keep cards small and focused. One card should test one thing — ideally recognizing or recalling a single word’s meaning in context. Cramming a whole definition, three synonyms, and a grammar note onto one card makes it slow to review and easy to fail for the wrong reasons.

Rating difficulty

After you try to recall the answer, most systems ask you to grade yourself — something like Again, Hard, Good, Easy. This is where the “ease” or difficulty rating comes in:

• Again — you couldn’t recall it. The card resets to a short interval and comes back soon.
• Hard — you got it, but it was a real struggle. The next interval grows only a little.
• Good — you recalled it with normal effort. The interval expands as expected.
• Easy — it was effortless. The interval jumps well ahead so you don’t waste reviews on it.

Be honest with these ratings. The whole schedule depends on them. If you tap Easy on everything to feel productive, you’ll push words out too far and forget them; if you tap Again out of perfectionism when you basically knew it, you’ll drown in needless reviews.

Intervals and the algorithm

Under the hood, an algorithm turns your ratings into a schedule. The long-standing classic is SM-2, the algorithm popularized by Anki and older SuperMemo versions. It keeps an “ease factor” for each card and multiplies the interval by it on each successful review, shrinking the ease when you struggle. It’s simple, transparent, and has served millions of learners well.

A more modern approach is FSRS (the Free Spaced Repetition Scheduler), which models memory more richly — tracking things like the estimated stability and retrievability of each memory — and adapts to your personal review history to target a retention rate you choose. You don’t need to understand the math to benefit; the practical upshot is that newer algorithms tend to schedule reviews more efficiently, giving you similar retention for fewer reviews. Either way, your job is the same: show up daily, recall honestly, rate honestly.

Download oqou on the App Store and you can turn any word you tap while reading into a review card, with the original sentence saved as context automatically.

Should you make your own flashcards or “mine” words from what you read?

Mine words from what you read. Pulling vocabulary out of real sentences you actually encountered — often called “sentence mining” — beats downloading someone else’s premade deck almost every time. The words are relevant to you, they arrive with built-in context, and each one carries a memory of where you met it, which makes it far easier to retain.

There are two broad ways to build a vocabulary deck, and they lead to very different results.

Premade decks: fast to start, weak to keep

Downloading a “top 2,000 words” deck feels efficient. Sometimes it’s a reasonable jump-start for absolute beginners who need a core of high-frequency words. But premade decks have real problems. The words arrive with no personal context, often with a single translation that ignores how the word actually behaves. You end up memorizing card-specific cues rather than the language — recognizing a word only in the exact format the deck presents it, then failing to spot it in the wild.

Mined words: slower to build, far stickier

When you mine a word from your own reading, you get several advantages at once:

• Relevance. You only add words you actually met, which skews your deck toward vocabulary that matters for the things you read.
• Context. The sentence comes along for free, and reviewing the word in that sentence is dramatically more effective than reviewing it alone.
• A memory anchor. “This is the word from that detective story” is a hook a premade card can never give you.
• Natural frequency weighting. Words you meet often in your reading get added often, so your deck automatically emphasizes what’s common in your world.

There’s also a compounding benefit that’s easy to miss. Because you meet a mined word in reading before you ever review it, and then keep meeting it in future reading, each SRS review reinforces something your reading is already touching. The deck and the reading feed each other. A premade deck sits off to the side of your actual language exposure; a mined deck is woven straight into it.

The one catch with mining is friction. If capturing a word means pausing, switching apps, typing out the word, finding a translation, and copying the sentence, you simply won’t do it — not for the dozens of words a good reading session throws at you. The trick is to make capture effortless. In oqou, tapping a word gives you an in-context meaning instantly and lets you save it with a single tap, sentence and all, so mining happens as a natural byproduct of reading rather than a chore that interrupts it.

Because context is doing so much of the work here, it’s worth understanding why it’s so powerful. We go deep on that in our guide to learning vocabulary in context — the short version is that meaning in a real language lives in usage, not in a one-word gloss, and cards that preserve usage teach you the word as it actually works.

How many new words per day is realistic?

For most learners, somewhere between five and twenty new words a day is a sustainable, realistic range. Beginners and busy people should aim low — even five a day adds up to well over a thousand a year. The number that matters isn’t your peak; it’s the number you can hit consistently for months without burning out.

The tempting mistake is to load up. Twenty, thirty, fifty new words a day — it feels ambitious and fast. But every new word you add today becomes a review you owe tomorrow, and the day after, and for weeks beyond. Front-load too hard and your daily review count balloons until a “quick” session takes an hour, you start dreading it, and then you quit. The people who succeed with spaced repetition are almost never the ones who added the most words per day. They’re the ones who never stopped.

A realistic way to set your number

• Start smaller than feels satisfying. Five to ten new words a day is plenty when you’re building the habit. You can always raise it.
• Watch your review load, not your new-card count. If your daily reviews are creeping past what you’ll actually do, lower the new words until it settles.
• Think in years, not weeks. Ten words a day is roughly 3,600 a year. That’s a serious vocabulary. Consistency, not intensity, gets you there.
• Let reading set the ceiling. If you’re mining from what you read, your natural intake is limited by how much you read — a healthy, self-regulating cap.

How big does your vocabulary actually need to get? That depends on your goal, and the honest answer surprises people — you need fewer words for functional fluency than you’d guess, but they need to be the right words, deeply known. We break the numbers down in how many words you need. The relevant point here: at ten well-chosen words a day, you reach meaningful milestones faster than almost any cramming sprint could sustain.

What are the most common spaced-repetition mistakes?

The most common mistakes are adding too many cards, making cards with no context, dishonest self-rating, and letting “leeches” — stubborn cards you fail over and over — pile up unaddressed. Each one quietly turns a powerful system into a daily chore you’ll eventually abandon. Almost all of them are easy to fix once you can name them.

Over-adding cards

This is the big one, and it deserves repeating because nearly everyone does it early. Adding is fun; reviewing is work. When you add faster than you can sustainably review, the debt compounds until the review queue is a wall. Discipline on intake is the single most protective habit in spaced repetition. When in doubt, add fewer.

Front-loading

Closely related: dumping a huge batch of new words in one enthusiastic session, then facing an avalanche of reviews for weeks. A steady trickle beats a flood. Spreading your additions across days keeps your review load smooth and predictable instead of spiky.

Cards with no context

A card that’s just word → translation is fragile. You memorize the pairing without learning the word’s behavior, and you fail to recognize it when it shows up conjugated, in a different sense, or in a real sentence. Always keep an example sentence on the card. If you mined the word from reading, you already have the perfect one.

Leeches

Leeches are cards you keep getting wrong no matter how many times they come up — they “leech” a disproportionate share of your review time. Every deck grows a few. The fix is not to keep grinding them. Instead:

• Reformulate the card. Often a leech is failing because the card is bad — ambiguous, overloaded, or missing context. Rewrite it with a clearer sentence.
• Add a mnemonic or a second example. Give the word another hook.
• Suspend or delete it. If a word keeps failing and isn’t important, let it go. You’ll meet it again in your reading, and next time it may stick on its own.

Dishonest or perfectionist rating

Rating every card Easy to feel fast, or every near-miss Again out of perfectionism, both corrupt the schedule. Rate by the actual effort of recall, not by how you wish you’d done. The algorithm is only as good as the honesty you feed it.

Treating reviews as the whole strategy

A subtler mistake: believing that flashcard reviews alone will make you fluent. They won’t. Spaced repetition is fantastic at moving known words into durable memory, but words come alive through use — reading them in context, hearing them, producing them. Reviews should support a life of reading and listening, not replace it. The learners who thrive treat their deck as a memory aid attached to a rich diet of real language, not as the main event.

How do you put it all together into a routine?

Build a simple daily loop: read something you enjoy, tap and save the words you don’t know with their sentences, add a small capped number of them as new cards, and clear your due reviews in a short session. Keep intake modest, rate honestly, prune leeches, and let the schedule run. Consistency over months is the whole game.

A concrete version of that routine might look like this:

1. Read for enjoyment first. Pick material slightly above your comfort level — hard enough to meet new words, easy enough to follow the story. Interest is what keeps you coming back.
2. Capture as you go. When you hit a word you don’t know, tap it, get the in-context meaning, and save it with its sentence. Don’t break your reading flow to over-analyze; just capture and continue.
3. Add a capped batch. Later, turn a limited number of saved words — say ten — into review cards. Leave the rest for another day.
4. Clear your reviews daily. Do your due cards in one short sitting. Recall first, then check, then rate honestly.
5. Prune weekly. Every so often, reformulate or retire your leeches and glance at your review load. Adjust your new-word rate if the queue is growing uncomfortable.

None of these steps is hard. The difficulty is entirely in doing them on the boring days — the days you’re tired, the days progress feels invisible. That’s where the design of your tools matters. The less friction between reading a word and reviewing it, the more likely you are to keep the loop alive for the months and years that actually build a vocabulary.

Get oqou on the App Store to keep that whole loop — read, tap, save, review — inside one app, on-device and private, so building vocabulary from real reading feels less like homework and more like just reading.

Frequently asked questions

How long until spaced repetition actually works?

You’ll notice individual words sticking within days, but the real payoff — a large, durable vocabulary — builds over months. Because the method deliberately spaces reviews out, it’s a slow-and-steady tool by design. Trust the process for a few weeks before judging it; the words that feel shaky now will feel automatic later.

Is spaced repetition better than just reading a lot?

They’re partners, not rivals. Reading widely exposes you to words in rich, varied context and is irreplaceable. Spaced repetition catches the words that would otherwise slip away between encounters and locks them in. The strongest approach combines both: read for exposure and enjoyment, then use spaced review to make the new words permanent.

Do I need a special app, or can I use paper flashcards?

Paper works and taught generations of learners, but it can’t schedule intervals for you — you’d have to track timing by hand, which quickly becomes unmanageable past a few dozen cards. A digital SRS handles scheduling automatically and scales to thousands of words, which is the whole reason the method is practical for serious vocabulary building.

What’s the difference between SM-2 and FSRS?

SM-2 is the classic, simple algorithm behind Anki and older SuperMemo — it adjusts each card’s interval using an “ease factor.” FSRS is a newer model that estimates the stability and retrievability of each memory and adapts to your personal history, generally scheduling reviews more efficiently. As a learner you don’t need to choose consciously; both aim to show you words right before you forget them.

Should I review vocabulary in both directions?

It depends on your goal. Recognition (target word → meaning) is enough for reading and listening, and it’s where most reading-based learners start. Production (meaning → target word) is harder and matters more if you want to speak and write fluently. If you’re building vocabulary from reading, recognition cards are a perfectly good foundation you can extend later.

Spaced repetition isn’t a trick or a shortcut — it’s just an honest way of working with how memory actually behaves instead of against it. Read things you love, capture the words that matter to you, review them at the right moments, and keep the intake modest enough to sustain. Do that for a few months and you’ll be surprised how many words have quietly become yours. If you’d like a reading-first way to build that habit, oqou is designed to make every step of it feel effortless.