Published August 4, 2026. Updated September 6, 2026

What order should you learn Japanese words in?

Learn the most common words first. A beginner who studies Japanese in frequency order meets the words that appear in nearly every sentence within the first few weeks, which means the study pays off the same day, in whatever anime, menu, or conversation you turn to next. Textbook and thematic orders sort words by topic instead, so you learn eleven kinds of fruit and the word for “post office” before you learn the words that hold sentences together.

Moji - Learn Japanese Daily for iPhone orders its whole catalog this way. All 8,237 entries sit in one deliberate sequence, starting from frequency data and then hand-edited so the front of the list actually teaches. This page explains where the ranks come from, why particles are pulled forward, and where frequency order stops being the right guide.

Why does frequency order beat textbook order?

A textbook chapter is organised for the writer’s convenience, not your exposure. Topic units (“at the restaurant”, “family members”, “colours”) make a satisfying table of contents, and they guarantee that a large share of what you learn in month one will not appear again for months. You learn 伯母 (おば, aunt) alongside 姉 (あね, older sister) because they belong to the same family unit, even though one of them is far more common than the other.

Frequency order inverts that. It sorts by how often a word actually shows up, so your first hundred cards are the words that recur constantly. The payoff is compounding: every common word you learn makes the next sentence you meet slightly more transparent, and because common words are common, that happens immediately rather than eventually.

Look at the 100 most common Japanese words at the bottom of this page and the pattern is hard to miss. Nine of the ten most frequent entries are particles. Twenty of the top hundred are. The rest are the plainest possible verbs and nouns: する (to do), ある (to be, to exist), 思う (おもう, to think), 人 (ひと, person), こと (thing). Nothing in that list belongs to a topic. It is the connective tissue of the language, and a topic-organised course has nowhere natural to put it.

Where do the frequency ranks come from?

Moji’s frequency ranks are derived from JPDB’s frequency data, matched to each entry on both word and reading so that homographs do not inherit each other’s ranks. 8,000 of the 8,237 entries carry a rank. Credit where it is due: JPDB did the corpus work, and the ranking underneath Moji’s order is theirs, not ours.

What Moji adds is the editing. The shipped study order is frequency-derived and then hand-edited, because a raw ranking is not a curriculum. A pure frequency list would open with the four or five most abstract function words in the language and expect you to make sense of them with no vocabulary to attach them to. The hand-editing exists to fix exactly that problem, and the clearest place to see it is the particles.

Why does Moji teach particles as vocabulary?

This is the part that makes Moji’s order different from a sorted spreadsheet.

Most courses treat particles as grammar. They get a chapter, a table, and an explanation of は versus が that beginners famously bounce off. Moji treats them as vocabulary: individual cards, learned and reviewed like any other word, with an example sentence doing the explaining. The particles guide walks the core set one sentence at a time.

That only works if the example sentence is readable. “Topic marker” is not a meaning you can picture, so a particle card carries almost no information on its own. The sentence is the card. Which means a particle cannot be introduced until the words its example needs are already known, and that constraint is what sets the position.

So the particles sit well behind their raw frequency:

The gap between column two and column three is the hand-editing. の is the single most common word in Japanese and it still waits until card 53, because the first fifty-odd cards are teaching kana and a starter set of concrete nouns and verbs, and without those there is no sentence to put の into.

Then each particle lands on an example built from words you already have:

• の (の, of, possessive), card 53: 私の猫。 “My cat.” 私 (わたし, I) is card 4 and 猫 (ねこ, cat) is card 20.
• は (read わ, topic marker), card 63: 夢はあります。 “I have a dream.” ある (to exist) is card 33. 夢 (ゆめ, dream) is the one word you have not met yet.
• が (が, subject marker), card 89: 水があります。 “There is water.” 水 (みず, water) is card 43, ある is card 33. Nothing here is new except が itself.
• か (か, question marker), card 109: 犬ですか。 “Is it a dog?” 犬 (いぬ, dog) is card 55 and です (is, am, are) is card 18.
• を (read お, direct object marker), card 129: 水を飲みます。 “I drink water.” 水 is card 43 and 飲む (のむ, to drink) is card 44, sitting side by side more than eighty cards before the particle that needs them both.

Read those five sentences in order and the particle is the only unknown in almost every one. You are not being taught a rule about を. You are being shown where it goes, in a sentence you can already read, which is a thing a beginner can absorb in a few seconds and then recognise everywhere.

It also puts は and が twenty-six cards apart with near-identical sentence frames, 夢はあります and 水があります, which is a gentler first encounter with that distinction than any explanation.

Two details worth noting. First, particles have readings that do not match their kana: は is read わ here and を is read お, which is a genuine early stumbling block, so both are annotated on the card. Second, the ordering is interleaved with the kana cards, since the catalog includes 208 kana-stage entries and a particle cannot be read before its own kana is. If you want the particles themselves explained rather than just placed, that is the Japanese particles guide.

Do the example sentences keep up?

Ordering the words is only half the job. An example sentence that introduces five new words to teach one is not comprehensible input, it is a wall.

So the example sentences for the first 600 words are written to a difficulty budget: past the opening stretch, each sentence uses at most one word the learner has not yet been taught at that point (katakana loanwords, which sound out their own meaning, are exempt). That is why 夢 stands out in the は example above. It is the budget being spent, deliberately, once. Holding that line across hundreds of sentences took a dedicated rewrite pass, and it is the subject of its own page: why example sentences are too hard.

Every one of the 8,237 entries has a curated example sentence with an English meaning. The budget applies to the front of the order, where it matters most.

What about JLPT levels?

JLPT level and frequency answer different questions. The level tells you what a test expects at each stage; frequency tells you what you are going to run into on Tuesday. They disagree more than you would expect, because the JLPT lists are a test curriculum rather than an exposure ranking, and plenty of extremely common words sit at the harder levels.

Moji’s position is that level should be a filter, not a sort order. If you are studying for a specific exam you can switch it on and work through the 775 N5 entries, or the 730 at N4, in frequency order within that level. If you are not, the level never gets in the way of the ordering. More on the exam-focused path: best JLPT N5 vocabulary app.

Where frequency order falls short

Three honest limitations.

Frequency data reflects whatever corpus produced it. A ranking built from one kind of Japanese will not perfectly describe another, and no frequency list is a neutral description of “the language”.

Frequency cannot tell you what you want to say. It is a ranking of what you will encounter, which is not the same as the vocabulary for introducing yourself, ordering food, or talking about your job. Those are needs a ranking has no way to see.

And raw frequency makes a bad first week. That is the whole reason Moji’s order is hand-edited rather than sorted: the top of a pure frequency list is abstract grammatical machinery, and handing it to someone on day one is how people conclude that Japanese is impossible. More on how the catalog is built and reviewed: how we curate Japanese vocabulary.

The 100 most common Japanese words

Word, reading, and meaning, in frequency order.

Note that this is the frequency ranking, not Moji’s study order. The study order takes these ranks as its starting point and then moves things, for the reasons above.

Try it

If a hand-edited frequency order sounds like the right way to spend your first months, Moji is free to start: the full comprehension path, flashcards with example sentences and mnemonics, costs nothing. Download Moji on the App Store And for how far the order has to run before Japanese feels usable, see how many Japanese words you need to be fluent.

Sources

• JPDB, Japanese frequency data used to derive Moji’s ranks: https://jpdb.io
• JMdict / EDRDG, the dictionary base for Moji’s entries: https://www.edrdg.org