Japanese verb groups: identify the group, and conjugation stops being scary
Every Japanese verb belongs to one of three groups. Once you can tell which group a verb is in, every conjugation — polite, negative, te-form, past — follows the same predictable rule. Here's how to sort them.
Japanese conjugation has a reputation for being hard. It isn't — it's just gated behind one prerequisite skill: knowing which of three groups a verb belongs to. Get that, and every conjugation you'll ever learn is a rule applied to a group, not a thing memorized per verb.
This article is the sorting step. We'll name the three groups, then give you the test for the one ambiguous case that causes nearly all the mistakes.
The three groups, named
Textbooks use different labels for the same three groups. Don't let the vocabulary confuse you — they're the same thing:
Group 1 — う-verbs / godan / "u-verbs". The big, default group. Dictionary forms end in a wide range of sounds: 飲む (nomu, drink), 書く (kaku, write), 話す (hanasu, speak), 買う (kau, buy), 待つ (matsu, wait).
Group 2 — る-verbs / ichidan / "ru-verbs". Always end in る, and crucially the sound before る is an /e/ or /i/ sound: 食べる (taberu, eat), 見る (miru, see), 寝る (neru, sleep).
Group 3 — irregular. Exactly two verbs: する (suru, do) and 来る (kuru, come). That's the entire group. Memorize the two and irregularity is handled.
So the only real sorting question is Group 1 vs Group 2 — and that's where the famous trap lives.
The trap: verbs that end in る but are Group 1
Here's the rule that feels safe but isn't: "ends in る → Group 2." It's a good first guess (most る-ending verbs really are Group 2), but a handful of common, everyday verbs end in -iru or -eru and are secretly Group 1. You have to learn these by heart:
帰る (kaeru, return home), 入る (hairu, enter), 走る (hashiru, run), 知る (shiru, know), 切る (kiru, cut), 要る (iru, need), しゃべる (shaberu, chat), 限る (kagiru, be limited).
Why it matters: 帰る conjugates like a Group 1 verb, so its te-form is 帰って (kaette), not 帰て. If you mis-sort it as Group 2, every single conjugation comes out wrong. These are high-frequency words, so the cost of mis-sorting is high.
There's no shortcut — this short list just has to be memorized. The good news: it's short, and once these ~8 are in your head, the "る → Group 2" guess is reliable for almost everything else.
Why the group decides everything: the ます-form
Take the polite ます-form (-masu) as the proof. Once you know the group, the rule is mechanical:
Group 1: change the final /u/ sound to /i/, then add ます. 飲む → 飲みます (nomu → nomimasu), 書く → 書きます (kaku → kakimasu), 待つ → 待ちます (matsu → machimasu), 話す → 話します (hanasu → hanashimasu).
Group 2: drop る, add ます. 食べる → 食べます (taberu → tabemasu), 見る → 見ます (miru → mimasu).
Group 3: memorize. する → します (shimasu), 来る → 来ます (kimasu).
Notice 待つ → 待ちます (tsu becomes chi, not ti) and 話す → 話します — these follow the Japanese sound chart, not a strict letter swap. That's the one subtlety in the Group 1 rule.
The same group powers every other form
The grouping you just learned is not a one-off for the ます-form. It's the same gate for nearly every conjugation in the language:
Plain negative (ない-form): 飲む → 飲まない (nomanai), 食べる → 食べない (tabenai).
Past plain (た-form): built on the same sound changes as the て-form — 飲む → 飲んだ (nonda), 食べる → 食べた (tabeta).
Potential ("can do"): 飲む → 飲める (nomeru), 食べる → 食べられる (taberareru).
Volitional ("let's"): 飲む → 飲もう (nomou), 食べる → 食べよう (tabeyou).
Every one of these is "identify the group, apply the rule." That's why the sorting skill is the real foundation — invest in telling Group 1 from Group 2 cleanly, and the rest of the conjugation system unlocks one rule at a time.
Written by
The Norolu Learning JP team
The editorial team behind Learning JP at Noroshi Inc., a small Japanese company in Mine, Yamaguchi. Every example, audio file, and etiquette note is selected and reviewed by the operator, one at a time.
Published:
Open the app
1,200+ situations and 10,000+ phrases. Pick one and try it on.