Reading a headline like "România are 19 milioane de locuitori" or a price tag marked "1.250,50 lei" requires two things English never prepared you for. First, Romanian swaps the punctuation: a period groups thousands (1.000.000) and a comma marks the decimal (3,14) — the exact mirror image of English. Second, the big number words mie, milion, miliard are real nouns that take de before whatever they count (milioane *de oameni*). Misread the punctuation in a contract or a statistic and you are off by a factor of a thousand — so this is a survival skill, not a nicety.
The punctuation swap — read this twice
Romanian and English use the same two symbols for grouping numbers, but assign them the opposite jobs:
| English | Romanian | |
|---|---|---|
| thousands separator | comma — 1,000,000 | period — 1.000.000 |
| decimal separator | point — 3.14 | comma — 3,14 |
| "one and a half thousand" | 1,500 | 1.500 |
| "three point five" | 3.5 | 3,5 |
So a Romanian shelf tag reading 3.000 means three thousand, not "three". And 3,5 kg means "three and a half kilos", not three thousand five hundred. This is the single most consequential difference on the page, because the symbols look familiar but lie to an English eye.
Mașina costă 25.000 de euro, scrie pe anunț.
The car costs 25,000 euros, it says in the ad.
Pâinea cântărește 0,5 kilograme.
The loaf weighs 0.5 kilos.
Thousands: o mie, două mii
A thousand is o mie — a feminine noun. Like any feminine noun it pluralizes (o mie → două mii) and, being feminine, takes două (never *doi) for "two thousand".
| Number | Romanian |
|---|---|
| 1.000 | o mie |
| 2.000 | două mii |
| 5.000 | cinci mii |
| 10.000 | zece mii |
| 100.000 | o sută de mii |
| 1.500 | o mie cinci sute |
Note 100.000 = o sută *de mii — "a hundred *of thousands" — because sută counts mii and twenty-plus counting puts de between them. The mii themselves then take de before the final noun: o sută de mii *de oameni* ("a hundred thousand people").
La protest au ieșit peste o sută de mii de oameni.
Over a hundred thousand people came out to the protest.
Apartamentul s-a vândut cu nouăzeci de mii de euro.
The flat sold for ninety thousand euros.
Millions and billions
Romanian keeps marching in units: un milion (a million, neuter) and un miliard (a billion, i.e. a thousand million, neuter). Note that miliard is the European "billion" — 10⁹ — which matches the modern English "billion"; Romanian has no ambiguity here, but the word itself is the false-friend-shaped miliard, not *bilion.
| Number | Romanian | Note |
|---|---|---|
| 1.000.000 | un milion | neuter |
| 2.000.000 | două milioane | plural milioane |
| 1.000.000.000 | un miliard | neuter; = English "billion" |
| 2.000.000.000 | două miliarde | plural miliarde |
Both are neuter, so "two million / two billion" use the feminine-shaped două (the neuter goes feminine in the plural — două milioane, două miliarde), exactly the agreement logic from the gender page.
Bugetul proiectului depășește două miliarde de lei.
The project's budget exceeds two billion lei.
Filmul a fost văzut de peste un milion de spectatori.
The film was watched by over a million viewers.
The de rule: mie, milion, miliard all take de
Because mie, milion, and miliard are nouns of quantity, they link to the counted noun with de — mii *de oameni, milioane **de lei, miliarde **de stele. This is the same twenty-plus *de you already know, and it's obligatory.
| Romanian | English |
|---|---|
| o mie de oameni | a thousand people |
| trei milioane de locuitori | three million inhabitants |
| cinci miliarde de euro | five billion euros |
| mii de mesaje | thousands of messages |
Orașul are aproape două milioane de locuitori.
The city has almost two million inhabitants.
Compania a investit zeci de milioane de lei în fabrică.
The company invested tens of millions of lei in the factory.
Reading decimals in statistics
In reports and news you constantly meet decimals — growth rates, averages, prices. Romanian reads the decimal comma as virgulă ("comma"): 3,5 is trei virgulă cinci ("three comma five"). (The full decimal-and-fraction system is on the fractions page; here is the statistics-reading essential.)
| Written | Read aloud |
|---|---|
| 3,5 | trei virgulă cinci |
| 2,75 | doi virgulă șaptezeci și cinci |
| 0,3 | zero virgulă trei |
Economia a crescut cu trei virgulă cinci la sută anul trecut.
The economy grew by three point five percent last year.
Rata inflației a ajuns la șapte virgulă doi la sută.
The inflation rate reached seven point two percent.
Percentages: la sută
A statistic almost always carries a percentage. Romanian says la sută ("per hundred") after the number: cincizeci la sută (50%), and the measured group links with din ("of"): cincizeci la sută *din populație. The formal noun for a percentage point is *procent.
Peste șaptezeci la sută din respondenți au fost de acord.
Over seventy percent of respondents agreed.
Prețurile au scăzut cu zece la sută față de anul trecut.
Prices fell by ten percent compared to last year.
Șomajul a urcat cu două procente în doar o lună.
Unemployment rose by two percentage points in just a month.
Common Mistakes
Reading a thousands-period as a decimal (the costliest error):
❌ '3.000 de lei' = three lei
Incorrect — the period groups thousands: 3.000 = three thousand lei.
✅ '3.000 de lei' = trei mii de lei
3,000 lei = three thousand lei
Writing a decimal with an English point:
❌ Inflația e de 3.5%.
Incorrect — decimals use a comma: 3,5%.
✅ Inflația e de 3,5%.
Inflation is 3.5%.
Dropping de after milioane / mii:
❌ trei milioane locuitori
Incorrect — these quantity nouns need 'de': trei milioane de locuitori.
✅ trei milioane de locuitori
three million inhabitants
Using masculine doi with the feminine mii or the neuter milioane:
❌ doi mii de lei / doi milioane
Incorrect — feminine/neuter take două: două mii, două milioane.
✅ două mii de lei / două milioane
two thousand lei / two million
Saying punct for the decimal, English-style:
❌ trei punct cinci la sută
Incorrect — the decimal is virgulă: trei virgulă cinci.
✅ trei virgulă cinci la sută
three point five percent
Key Takeaways
- Romanian swaps the punctuation: period for thousands (1.000.000), comma for decimals (3,14) — the reverse of English. The thousands-period is silent.
- Mie, milion, miliard are nouns: they pluralize (mii, milioane, miliarde), take de before the counted noun (milioane *de oameni), and use feminine *două for "two" (două mii, două milioane).
- Miliard = English "billion" (10⁹); don't reach for *bilion.
- Read decimals with virgulă (3,5 = trei virgulă cinci) and percentages with la sută (cincizeci la sută din…).
Now practice Romanian
Reading grammar gets you part of the way. The exercises are where it sticks — free, no signup needed.
Start learning Romanian→Related Topics
- Cardinal Numbers 20 and AboveA1 — The tens (douăzeci…nouăzeci), compound numbers built with 'și' (douăzeci și unu = 21), hundreds and thousands, and the rule that defines Romanian counting above twenty: from 20 up, the number connects to its noun with 'de'.
- Fractions, Decimals, and PercentagesB1 — Romanian reads decimals with a comma (3,14 = trei virgulă paisprezece), builds fractions beyond half/quarter from ordinal-derived -ime nouns (treime, pătrime), and expresses percentages with 'la sută'. This page flags the comma-not-point trap that catches every English speaker.
- Number-Noun Agreement and 'de'A2 — Only 1 and 2 inflect for gender in Romanian (un/o, doi/două) — but they keep agreeing even inside huge compounds (treizeci și două de cărți), and the neuter counts with the feminine form. This page also consolidates the 'de' threshold at twenty.
- Telling Dates and TimeA2 — Dates use plain cardinals plus a month (pe 5 martie) — except the 1st, which is the special ordinal 'întâi'; clock time uses 'și' for minutes past the hour (trei și zece) and 'fără' ('without') for minutes to the hour (patru fără cinci).
- Quantifiers (mult, puțin, tot, câțiva)B1 — Romanian quantifiers — mult/puțin (much/little), destul (enough), tot (all), câțiva (a few), atât (so much) — with their agreement as determiners versus their invariable adverbial use, the trap that makes one word run on two grammars.