Japanese Pitch Accent: The Complete Guide for Learners (2026)

Japanese pitch accent is the feature of pronunciation that most textbooks skip entirely — and the feature that most separates advanced learners from native-sounding speakers. Unlike Chinese tones, it is not a tone on every syllable. Unlike English stress, it is not about volume. It is a system of high and low pitches that follow predictable patterns, and once you understand those patterns, training your ear and your production becomes surprisingly methodical.

This guide covers everything: what pitch accent actually is, how the Tokyo standard system works, the four accent patterns with real examples, why it matters for comprehension, the minimal pairs where it changes meaning, and the tools and practice methods that actually move the needle.

What is Japanese pitch accent?

Japanese is a pitch-accent language. Each word in standard Tokyo Japanese is assigned a pattern of high (H) and low (L) pitches across its morae (the rhythmic units, roughly equivalent to each hiragana character). The key feature of the system is the downstep — a position in the word where pitch drops from high to low and does not rise again within that word.

The pitch always starts low on mora 1, rises to high on mora 2 (with one exception — the atamadaka pattern), and then either stays high or drops at a specific position. Where that drop falls — or whether it falls at all — defines the accent pattern.

Crucially, Japanese pitch accent is not like English stress. English stress adds loudness, length, and vowel clarity to one syllable. Japanese pitch accent is purely about the melody — the high-low contour. A low-pitched mora in Japanese is not quieter or shorter; it is simply lower in pitch.

Morae: the unit pitch accent is built on

Before the patterns make sense, you need to be clear about morae. A mora is slightly different from a syllable. In Japanese, each hiragana character counts as one mora, including:

• Regular kana: か, き, く, け, こ — each = 1 mora
• Long vowels: おかあさん = o-ka-a-sa-n = 5 morae (the extra あ counts)
• Double consonants: きって = ki-t-te = 3 morae (the っ counts)
• ん: ほん = ho-n = 2 morae
• Combined kana: きょ counts as 1 mora (not 2)

Pitch accent notation numbers morae from 1. So “accent on mora 2” means the downstep happens after the second mora.

The four Tokyo pitch accent patterns

Standard Tokyo Japanese (and the Japanese taught in NHK broadcasts) uses four distinct accent patterns. Every noun, verb, and adjective falls into one of these categories.

1. Heiban 平板 (flat) — accent type 0

Heiban words start low on mora 1, rise to high on mora 2, and stay high for all remaining morae. There is no downstep within the word. However, any particle that follows will be pronounced low.

2. Atamadaka 頭高 (head-high) — accent type 1

The word starts high on mora 1 and immediately drops to low on mora 2, staying low for the rest of the word (and any following particles). This is the only pattern where mora 1 is high.

3. Nakadaka 中高 (middle-high) — accent types 2, 3, 4…

The pitch rises to high after mora 1 and then drops somewhere in the middle of the word — not at the very end. The drop position gives the sub-type (accent 2 = drop after mora 2, accent 3 = after mora 3, etc.).

4. Odaka 尾高 (tail-high) — accent on final mora

Odaka words rise to high and stay high through the final mora of the word. The drop only happens on the first particle that follows. Without a particle, odaka and heiban sound identical — the difference only emerges when a particle is attached.

Minimal pairs: when pitch accent changes meaning

The practical reason to care about pitch accent is not just sounding native — it is avoiding genuine misunderstandings. Japanese has many homophone pairs where the only distinguishing feature is the accent pattern. In context, listeners can usually resolve these, but in isolation or in fast speech, wrong pitch can cause real confusion.

Sentence-level pitch: compound words and particles

Pitch accent does not operate only on isolated words. In natural speech, words combine into longer phrases, and the pitch pattern of the whole phrase is determined by rules about how individual word accents interact.

The key principle is the downstep is preserved but shifted in compound nouns. When two nouns combine, the accent of the second element often dominates, but the rules vary and are complex enough that even native speakers vary. For learners, the most important sentence-level rule is:

• After a downstep, pitch stays low for the rest of the phrase (including particles).
• Heiban words have no downstep, so particles following them also start high before dropping.
• Odaka words drop on the first particle — so 橋が is L-H-L (the が is low).

Regional variation: not just Tokyo

The Tokyo (Yamanote) dialect pitch accent system described above is the national broadcast standard, but Japan has significant regional variation:

• Kyoto/Osaka (Kansai-ben): A completely different pitch accent system with its own four-pattern structure. The same words often have opposite pitches to Tokyo. 雨 in Osaka is L-H (vs H-L in Tokyo).
• Tohoku: Largely unaccented or uses very different patterns.
• Kagoshima: Two-pattern system, very different from Tokyo.
• Kyushu: Mixed patterns depending on prefecture.

For most learners, mastering Tokyo/standard pitch accent first is the right approach. You can acquire regional patterns through exposure later.

Why pitch accent matters for listening comprehension

Even if you are not focused on your own production, understanding pitch accent dramatically improves your listening ability. Native speakers segment spoken Japanese into words partly using pitch contour — the pitch pattern is a boundary cue that tells listeners where one word ends and another begins. Learners who have no pitch awareness often struggle with word segmentation in fast speech.

Once you internalise the four patterns, you will start to automatically group morae into words as you listen, which makes audio-heavy study (podcasts, drama, anime) significantly more effective.

How to practise pitch accent: step-by-step method

Pitch accent is best trained through a combination of audio exposure, explicit pattern lookup, and active shadowing. Here is the most efficient progression:

Step 1 — Learn the four patterns conceptually

Before touching audio, be able to describe each pattern in your own words and draw the H/L contour for a 4-mora word in each category. This gives your brain a framework to file incoming audio into.

Step 2 — Look up and annotate new vocabulary

Every new word you add to your SRS or vocabulary list should include the accent number. Use Prosody Tutor (Suzuki-kun) or the pitch accent field in OJAD. Many Anki Japanese decks (Core 2k/6k, JPDB) include pitch accent data. Mark the accent visually — e.g., write か↘き (downstep after mora 1 = atamadaka).

Step 3 — Listen before you speak

For each new word, find a native pronunciation on Forvo orNHK Web Dictionary and listen at least three times before attempting to produce it. The goal is to have the correct pitch pattern in your auditory memory before you build a production habit.

Step 4 — Shadow with pitch-annotated scripts

Shadowing means repeating audio in real time, slightly behind the speaker. Use scripts with pitch accent notation (Dogen's course, Pitch Accent for Japanese Learners by Tofugu, or Accent Annotator browser extensions). Shadow at 0.8× speed first, then at 1× speed, focusing on matching the melodic contour, not just the sounds.

Step 5 — Record and compare

Record yourself producing words and short sentences. Compare your audio to a native recording using a spectrogram tool (Praat is free) or simply by ear. The goal is not perfection on first attempt — it is identifying systematic errors (e.g., “I always make compound nouns heiban”) and targeting those.

Common beginner pitch accent mistakes

Best tools for learning pitch accent