Nouns Whose Gender Contradicts Their Ending

The defaults on feminine endings are good ones: ־ָה and ־ֶת mean feminine, a final consonant means masculine. They are also broken in both directions by words you use every day, and because Hebrew is written unpointed, the adjective is often the only evidence on the page of which way a given noun goes. This page is the two lists.

Masculine nouns that end in ה

Most of this list is not really an exception, and seeing why makes it far easier to hold. Unpointed ה hides two different endings: the feminine marker ־ָה of דירה and שאלה, and a plain ־ֶה that was never a gender marker at all.

The first four are sadé, mikré, ma'asé, makhané — stressed on the final ־ֶה, and audibly not a ־ָה noun, which is stressed the same way but sounds -á. Only the last row is a genuine exception: לַיְלָה really does have the feminine ־ָה, really is masculine, and is stressed on the stem — láyla, not layl-á.

Three of those nouns take a ־ות plural anyway — שדות, מחנות, לילות — and so does כִּיסֵּא kisé "chair", which is masculine too and pluralises כִּיסְּאוֹת. A ־ות plural is a fact about the individual noun and says nothing about its gender, which is why לילות still takes masculine adjectives. That mismatch has its own page, plurals that lie about gender.

Feminine nouns with no marker at all

This is the longer list and the more damaging one, because every word on it is ordinary A1 or A2 vocabulary that a learner has been treating as masculine for months.

Be honest with yourself about this list: nothing predicts it. There is no sound, no shape and no meaning that these words share and that ספר, שולחן and כביש lack. They are feminine because they have been feminine since Biblical Hebrew, and the only way in is to learn each one inside a phrase — דרך ארוכה, אבן כבדה — rather than as a bare word.

The nouns that really are both

A few common words carry a זכר ונקבה tag in the dictionaries, and it is worth knowing which they are so that you do not "correct" a native speaker. Milog tags סַכִּין, שֶׁמֶשׁ, מַטְבֵּעַ, רוּחַ, לָשׁוֹן, כּוֹס and even דֶּרֶךְ this way; Pealim lists סכין as masculine.

But the tag mostly records historical variation rather than a live coin toss, and modern usage is far more settled than the label suggests. A Hebrew Wikipedia search on 21 August 2026 — the regex form of insource:, with a trailing space on each probe — returns דרך ארוכה 311 times against a single דרך ארוך, and that one hit is תוואי הדרך ארוך יותר, where ארוך agrees with תוואי and not with דרך at all. The real score is 311 to nothing. Feminine wins the same way on רוח קרה, לשון עשירה, כוס מלאה and השמש שקעה, none of which has a masculine counterpart in the corpus at all; he.wiktionary's tag for שמש says as much in words, נקבה בעיקר, גם זכר.

סכין is the real one. The same probe gives סכין חדה 26 against סכין חד 3, and all three masculine hits are genuine — סכין חד לחיתוך, סכין חד וחלק, בסכין חד. Both genders are alive, and you will hear both.

מטבע went the other way. The dictionary tag is real, but modern usage has settled on the masculine: מטבע זר 74 against מטבע זרה 0, המטבע הרשמי 66 against המטבע הרשמית 0. The feminine survives mainly on the physical coin in older writing. A זכר ונקבה tag tells you a noun once wavered; it does not tell you which way it came down.

The body parts that come in pairs

יָד, רֶגֶל, עַיִן, אוֹזֶן, שֵׁן, בֶּרֶךְ, כָּתֵף are all feminine and none of them is marked. They belong to the list above, but they also form a class you can actually predict, which is worth having on its own — the whole of it is on paired body parts, cities and countries. Their dual plurals (ידיים, עיניים, ברכיים) are a separate matter, treated on the dual.

Common Mistakes

1. Reading a final ה as the feminine marker. The commonest error on this page, and the fix is to hear the vowel: ma'asé has ־ֶה, dirá has ־ָה.

2. Letting one wrong guess run through the whole clause. Gender is not decided once per word — it is spent on everything that agrees, so a single mis-filed noun produces two or three visible errors.

3. Trusting a final consonant on a word from the feminine list. The default said "assume masculine", and assuming was right — until you met the noun and learned otherwise.

4. Correcting a Hebrew speaker who says הסכין הזה. They are not wrong, and neither is הסכין הזאת. When a word is genuinely two-gendered, pick one and be consistent inside your own sentence — that consistency is the only part that is not optional.

Key Takeaways

• שדה, מקרה, מעשה, מחנה are masculine and are not really exceptions: their ending is ־ֶה, which was never a feminine marker. לילה is the genuine one — a real ־ָה noun that is masculine, and stressed láyla.
• כיסא is masculine too, and its ־ות plural (כיסאות) proves nothing about gender — nor does לילות, שדות or מחנות.
• אבן, דרך, ארץ, עיר, נפש, בטן, אצבע are feminine with no marker at all, and nothing about them predicts it. Learn each inside a phrase: דרך ארוכה, אבן כבדה.
• The segolate shape is silent about gender — דרך, אבן, נפש and פעם are feminine segolates; ספר, בוקר, ילד and כלב are masculine ones.
• A זכר ונקבה dictionary tag usually records older variation, not a live choice: דרך, רוח, לשון, כוס and שמש take feminine agreement in modern usage, and מטבע has settled the other way, masculine. סכין is the one where both genders are genuinely current.
• The paired body parts — יד, רגל, עין, אוזן, שן, ברך, כתף — are all feminine and unmarked, and they form a predictable class of their own.