Spanish verb flashcards, built around conjugation

Here is a card that teaches almost nothing:

You can get that card right 40 times and still stall in a conversation, because Spanish almost never asks you for hablar. It asks for hablé, hablaba, hablo, hablamos, and it asks under time pressure with a person sitting opposite you.

So the card is testing a dictionary lookup, and the dictionary is the one part of this you will never need to do out loud.

Here is the card I'd build instead:

One hole, one form that fits it. To fill it you have to work out that this is a completed past action, that the subject is yo, and that ir goes irregular in the preterite. The card tests all of that in one go, because a sentence with a gap has exactly one correct shape.

That's the whole argument of this page, and I'll spend the rest of it on the specifics: which tenses, which verbs, in what order, and what any of it costs.

What the card should look like

A gapped sentence is called a cloze. Wilson Taylor published the procedure in Journalism Quarterly in 1953 as a readability measure, deleting words and counting how many readers could restore them. The name comes from closure: a mind shown an incomplete pattern moves to complete it.

For verbs specifically, the gap does work no word pair can do. A conjugated form carries person, number, tense and mood bundled into one ending. Strip the form out of a sentence and every one of those has to be reconstructed to fill the hole.

Four rules make the difference between a card that trains speech and a card that trains recognition.

Put the answer inside a sentence, never alone. fui on its own is ambiguous: it's the preterite of both ir and ser, which is a genuinely useful thing to know and an impossible thing to test without context. Ayer fui al médico resolves it. So does Fui profesor durante diez años, which is the other verb entirely.

Label the form on the back, in plain English. "preterite of ir" tells you what you just retrieved and lets the same verb hold several cards without them blurring into each other. A back that says only fui leaves you guessing which of the two verbs you just practised.

Produce it before you look. Say the form out loud, or type it. A spoken guess can't be quietly upgraded to a pass after the answer appears, which is the entire failure mode of flip-and-judge study. The rubric with the edge cases is in the grading post.

Get audio on the sentence if you can. Spanish verb endings are where the information lives and they are unstressed, fast and easy to mishear. Hablo and habló differ by one stressed syllable and about six months of confusion.

Which tenses, in which order

This is the decision that matters most on the whole page, and most guides duck it by listing every tense in a grid.

The order I'd defend: present, preterite, imperfect. Then ir a plus an infinitive for the future. Everything else waits.

I want to be careful here, because there's a number I can't give you. I looked for a corpus study putting a clean percentage on how much of real Spanish speech those tenses cover, and I didn't find one I could verify to a primary source. Plenty of pages state a figure. I'm not going to repeat one I can't check, so take the ordering as a defensible teaching argument rather than a measured claim.

The argument goes like this. The present does double duty in Spanish, covering habitual actions and, in speech, plenty of near-future ones. The preterite and the imperfect between them handle the entire past, and the choice between them is the single hardest grammar decision a Spanish learner makes, which is a reason to start early rather than late.

The periphrastic future, voy a comer, is almost free. You already conjugate ir. It covers most of what the simple future does in ordinary speech, and learners routinely reach for it over comeré.

What to defer, and why:

• The simple future and the conditional. Both regular enough to pick up quickly later, and both largely covered in speech by constructions you'll already own.
• The compound tenses. He comido, había comido. These are one auxiliary plus one participle, so they cost you one pattern once haber is solid, rather than a whole set of endings.
• The subjunctive. Unavoidable in Spanish, and the thing beginners start too early. It's defined by contrast with the indicative, so drilling it while the indicative is still wobbly means drilling a contrast with one soft half.

That last one is a real opinion and some teachers will disagree with it. I'd rather commit to it than give you a grid.

A starter deck you can build today

Here's a concrete set. It's ordered on teaching value rather than raw frequency, and the ordering principle is the interesting part: related forms are deliberately kept apart.

That comes out of the semantic clustering research. Thomas Tinkham found in 1993 and again in 1997 that learners taught semantically related words together needed significantly more trials to learn them; Rob Waring replicated it in 1997 with Japanese-to-artificial word pairs. The mechanism is interference. Present a learner with a tight set at once and they build one fuzzy node and then guess inside it, which is exactly what ser and estar feel like when they've been taught as a matched pair on a Tuesday afternoon.

So ser and estar go in different weeks. So do saber and conocer, and por and para if you're carding those. Contrast them later, once each is independently solid.

One note on week 7, because it's the kind of fact that changes how a tense feels before you start it: the Spanish imperfect has three irregular verbs. Ir, ser, ver. That's the list. Learners who've been braced for another wall of exceptions tend to find that genuinely cheering.

Block the introduction, mix the review

Here's where I can give you a real number, because somebody ran exactly this experiment on exactly these tenses.

Pan, Tajran, Lovelett, Osuna and Rickard taught adult learners to conjugate the Spanish preterite and imperfect across four experiments, published in the Journal of Educational Psychology in 2019. One group blocked, doing one tense at a time. The other interleaved, alternating between them.

In the single-session experiments, interleaving lost or drew. Experiment 1 gave "numerically higher performance in the blocked group"; Experiment 2 came out equivalent.

Then they spread training across two weekly sessions, with a delayed test a week after the second. The blocked group learned one tense per session; the interleaved group did both per session, same total material.

Pan et al. 2019, Experiment 4 · conjugation accuracy one week later

0

percent correct for the group that studied one tense at a time

0

percent correct for the group that alternated between the two tenses, d = 0.79

Now the part that matters more than the headline. During training, the blocked group performed better. They also rated their own learning higher, in both sessions, while the interleaved group rated the task as harder.

So the schedule that felt like it was working was the one that lost by 19 points a week later. That's the same shape as the testing effect and desirable difficulty: the method that feels worse in the room is the one that holds.

Two honest caveats, and the second one costs me something.

First, the win needed multiple sessions. Interleaving from the first minute of a single sitting did nothing, which is why the deck above blocks each new pattern on introduction and only mixes afterwards.

Second, Pan's own General Discussion says this, mid-victory-lap: "language is not a single capacity but a collection of many skill subdomains, and interleaving's benefits are likely to vary by task type; for instance, it has not shown a benefit for learning vocabulary."

Which is correct, and I've written about it at length. Brunmair and Richter's 2019 meta-analysis found interleaving actively hurts for word learning, at g = -0.39. I got that wrong on this site once and corrected it in public.

Verb conjugation is the case where interleaving is evidenced, because it's rule discrimination rather than word learning. Don't carry the result across to your vocabulary deck.

The form explosion, and how to survive it

Spanish verbs have a lot of forms. Six persons across a dozen-odd tenses and moods, which is where verb decks go to die: somebody makes a card per form per verb, hits four figures inside a fortnight, and quits.

The Anki manual puts the arithmetic plainly. "If you are consistently learning 20 new cards a day, you can expect your daily reviews to be roughly about 200 cards/day." Every card you add today is still generating reviews in six months.

So be ruthless about what earns a card.

A regular ending is a pattern, not 6 facts. One card with a worked example beats six cards drilling the same rule's output. You're learning an operation you apply, and applying it is what the cloze sentences already make you do.

An irregular form is a fact, and facts need cards. Hice, tuve, dije, supe. These don't fall out of any rule, so they earn individual cards with individual sentences.

Card the forms that beat you, not the forms that exist. This is the one that actually controls deck size. If you never miss the nosotros preterite, it doesn't need a card, whatever the conjugation table says. The best source for what beats you is your own speech: the verb you fumbled in a lesson last Tuesday is worth more than the next row of a conjugation chart.

Add fewer new cards than you want to. Five a day is plenty and compounds faster than you'd expect. Ten sounds modest and turns into a pile you didn't sign up for. More on the arithmetic in how many cards a day.

What each option costs

Not one of the three pages currently ranking for this search states a single price. FluentU lists 16 apps and prices none of them. FlashRecall says "start free, then scale up" with no tier amounts. Flashcardo has a "Buy Now" on an ebook with no figure attached. All three checked on 21 September 2026.

That seems like a bad deal for someone deciding where to put money, so here are real numbers, each one read off the company's own page today.

A few things worth saying plainly about that table.

Anki's free tier isn't a limited version of something. The desktop app is the full program, syncing costs nothing, and the iOS price is a one-time purchase that funds development rather than a subscription. If you'll enjoy making your own conjugation cards, Anki plus an afternoon is better than anything you can pay for, and I've argued that at length elsewhere. Turn FSRS on at the bottom of the deck options page while you're there, because it's opt-in and the manual is explicit that the default is still the 1987 SM-2 algorithm.

What paid apps sell you is the labour and the audio. Not a better scheduler: FSRS is free and open, and Anki runs it. That's the honest frame for every price in that table, including mine.

And a warning on ready-made conjugation decks, free or paid. A deck of 2,000 forms assembled by someone else is a deck of 2,000 forms you haven't met, which is the download-a-giant-deck failure mode. The tradeoff between premade and your own cards is real in both directions, and for verbs specifically it tilts hard toward the forms you've personally fumbled.

Which is the thing I built. Verbamor takes a lesson you recorded, finds the words and the verb forms you actually used or struggled with, and makes the cards: a gapped sentence, spoken audio in a native accent, and a plain label naming the form, "preterite of ir" rather than a dictionary headword. There are two recordings on every card, one with the answer replaced by a tone for the recall side and a complete read on the back. It covers five languages, it's iOS only, and it will not teach you your first thousand words, because it only holds what you've already met. You don't need a tutor for it (the Translate tab makes cards from words you type), though it's at its best mining something you actually said.

Questions people actually ask

---

Sources

Every load-bearing claim Verbamor makes is traced to its paper on the research page.