Vocabulary size

How many words do you need for A1, A2, B1, B2 and C1?

There is no official number: the CEFR describes vocabulary in words, not counts. The widely cited research figures, from learners who passed Cambridge English exams, are under 1,500 words at A1, 1,500–2,500 at A2, 2,750–3,250 at B1, 3,250–3,750 at B2 and 3,750–4,500 at C1. They count only the 5,000 most frequent words, so a C1 learner's whole vocabulary is larger.

By the Readlang Mobile team · Updated 1 November 2026

Is there an official number of words for each CEFR level?

No. The Common European Framework of Reference for Languages (CEFR) describes each level by what a learner can do, and its descriptors were deliberately kept “holistic”, leaving detailed vocabulary lists to specifications for particular languages (Council of Europe, 2001, p. 30). The Framework itself suggests that users “may wish to consider … what size of vocabulary (i.e. the number of words and fixed expressions) the learner will need” (p. 150), but it does not give the number.

What it does give is a vocabulary range scale in words. A1 is “a basic vocabulary repertoire of isolated words and phrases related to particular concrete situations”; B1 is “sufficient vocabulary to express him/herself with some circumlocutions on most topics pertinent to his/her everyday life”; C2 is “a very good command of a very broad lexical repertoire including idiomatic expressions and colloquialisms” (Council of Europe, 2001, p. 112, as reproduced in Milton, 2010).

Numbers come from two other places:

• Syllabus lists. Before the CEFR, the Council of Europe's specifications for what are now A2 (Waystage) and B1 (Threshold, Niveau Seuil) came with word lists of about 1,000 and about 2,000 words (Milton, 2010). Exam boards still publish lists: the word list for the Goethe-Zertifikat B1 in German covers “circa 2400 lexikalische Einheiten” that candidates should know, at least receptively (Goethe-Institut et al., 2016). These lists say what a course or exam expects, not what learners at that level actually know.
• Measurements of learners. Researchers have tested the vocabulary of learners already placed at a level. These are the figures in the next section, and the ones behind most numbers you will find online.

How many words do learners know at each CEFR level?

The most-cited figures come from Paul Meara and James Milton, who tested students taking and passing Cambridge English exams at every CEFR level with their X_Lex vocabulary test (Meara & Milton, 2003, reported in Milton, 2010).

Two things about the test shape these numbers. First, X_Lex is a yes/no checklist of real and invented words from the 5,000 most frequent lemmas; ticks on invented words lower the score, to correct for overclaiming (Milton, 2010). Second, the test cannot score above 5,000, and the authors note that learners also know words outside that range, so total vocabulary is underestimated.

The unit is the lemma: a base form with its most regular inflections, such as walk, walks, walked and walking, but not derived words like walker (Milton, 2010). Figures in word families, the unit of most reading research, are not directly comparable; our guide to what counts as a word and our page on how many words the average person knows explain the difference.

How many words do you need for A1 and A2?

For English, under 1,500 of the 5,000 most frequent words at A1 and 1,500–2,500 at A2, on the Cambridge exam bands above (Milton, 2010). Across the languages tested, Milton concluded that “something like 2000 words, of the most frequent 5000, might be needed to attain A2 level”. Groups of learners placed at these levels by their teachers averaged:

• A1: 1,477 for learners of English in Greece, 894 and 1,125 for learners of French in Spain and in Greece, and 1,492 for learners of Greek.
• A2: 2,156, 1,700, 1,756 and 2,237 in the same four groups.

For comparison, the old Council of Europe word list for Waystage, today's A2, had about 1,000 words (Milton, 2010). A list says what a course should cover; tested learners at A2 usually recognise more than that.

How many words do you need for B1 compared with B2?

For English, about 2,750–3,250 of the 5,000 most frequent words at B1 and 3,250–3,750 at B2, a difference of roughly 500 (Milton, 2010). Milton put the threshold for independence at about 3,000: “approximately 3000 words, of the most frequent 5000 words, might be needed to progress beyond the elementary A1 and A2 levels and achieve any kind of independence in English and Greek”.

In real groups the step from B1 to B2 can be small. Learners of English in Hungary averaged 3,135 at B1 and 3,668 at B2; in Greece, 3,263 and 3,304. For German, the Goethe-Zertifikat B1 word list has about 2,400 entries (Goethe-Institut et al., 2016); that is a syllabus, not a test result, and it leaves out compounds and derived words that can be worked out from words on the list.

How many words do you need for C1?

At least 3,750–4,500 of the 5,000 most frequent English words on the Cambridge exam bands, which means recognising almost all of the most frequent 5,000 (Milton, 2010). Learners of English at C1 averaged 4,340 in Hungary and 3,690 in Greece. Because the test stops at 5,000, it cannot say how many words a C1 learner knows in total. Milton argues that scores of 4,000 or more imply that “learners' overall vocabularies must be very large indeed”, consistent with Nation's estimate that 8,000–9,000 word families are needed to read novels and newspapers with 98% of the words known (Nation, 2006).

So the C1 figures of 8,000–10,000 words that circulate online are of the same order as Nation's reading estimate, but they are usually given without a source or a unit. The research behind them measures something else: recognition of the most frequent 5,000 lemmas, plus coverage estimates in word families. Our hub page explains how many words you need to read a novel.

Is it the same number in every language?

No. Milton and Alexiou tested learners of English, French and Greek with versions of X_Lex built the same way from each language's 5,000 most frequent lemmas, with levels set by the learners' teachers (Milton & Alexiou, 2009; summarised in Milton, 2010).

At every level the learners of French knew fewer of their language's frequent words than the learners of English; at every level with data for Greek (A1 to B2), the learners of Greek knew more. Milton offers two explanations: the most frequent French words may cover more of a typical text, so fewer are needed for the same tasks, or the corpora behind the three tests were built differently. He does not decide between them, and the samples are small.

Word counts also differ by language for a mechanical reason. Over the same 5,000 running words, the median book in our library uses 1,311 different word forms in English and 1,663 in Spanish, mostly because Spanish verbs and adjectives have more endings (our counts of 117 books). A figure for one language does not transfer to another, and a figure in word forms is not a figure in lemmas.

How many different words do books at each level use?

We counted the public-domain books in the Readlang Mobile library, each labelled with a CEFR level by the app's editors when it was added. The books are original texts, mostly classics, not graded readers, and the counts are word forms: walk and walked count as two, and names count too. They cannot be set against the X_Lex figures above, which are lemmas known by learners; they show how much vocabulary the books themselves demand.

The forms needed for 98% of the text rise from 2,532 at A1 to 10,848 at C1, but much of that rise is length: the median C1 book runs to 192,630 words, against 79,464 at B1, and longer books keep adding rare words. Over equal stretches of 5,000 words the differences shrink. Only the A1 median, from Alice's Adventures in Wonderland and Grimms' Fairy Tales, stands clearly apart, and even then two B2 non-fiction books, Descartes's Philosophical Works and Adam Smith's The Wealth of Nations, use as few forms per 5,000 words as Alice.

Only in German does the median rise steadily from level to level. In English, Spanish, French and Italian it moves up and down. The reason is that a level label on an original book reflects sentence length, style, subject and age of the language as well as vocabulary, and it is a judgement, not a measurement. Before you start a book, count the unknown words on one page; the two-minute check in our first-week plan shows how. Every book with its counts is listed on how many different words are in a book.

How can you check which level your vocabulary matches?

The bands on this page are X_Lex scores, out of the 5,000 most frequent lemmas. For English, the common free tests measure different things: Nation's Vocabulary Size Test reports word families, testyourvocab reports dictionary headwords, and LexTALE gives a percentage that tracks general proficiency. None of them uses the X_Lex scale, so their results cannot be set directly against the bands on this page. Our comparison of vocabulary size tests covers what each one measures, how accurate it is and which to take.

What this means for learners

• For A2, aim at the most frequent 2,000 words. That is roughly what learners at A2 recognise in every language Milton tested, and they are the words that make up most of any text.
• For B1, about 3,000 of the most frequent 5,000. This is the point Milton links to the first real independence in English and Greek; learners of French reached B1 with fewer (about 2,200–2,400).
• For C1, nearly all of the most frequent 5,000, and many more beyond. Reading unsimplified novels in comfort takes more again: about 8,000–9,000 word families (Nation, 2006).
• Expect it to take time. Regular reading and spaced review add the rest.
• Compare numbers only in the same unit and language. A count in lemmas, one in word families and one in word forms can describe the same learner, and the same level needs different counts in French and in English.

References

1. Council of Europe (2001). Common European Framework of Reference for Languages: Learning, teaching, assessment. Cambridge University Press. Pages 30, 112 and 150 as quoted in Milton (2010). coe.int
2. Goethe-Institut, ÖSD & Universität Freiburg/Schweiz (2016). Goethe-Zertifikat B1: Wortliste. Vorwort, pp. 5–6. goethe.de (PDF), checked 31 October 2026.
3. Meara, P., & Milton, J. (2003). X_Lex, the Swansea Levels Test. Newbury: Express.
4. Milton, J. (2010). The development of vocabulary breadth across the CEFR levels. In I. Bartning, M. Martin & I. Vedder (Eds.), Communicative proficiency and linguistic development: Intersections between SLA and language testing research (EUROSLA Monographs Series 1, pp. 211–232). eurosla.org (PDF)
5. Milton, J., & Alexiou, T. (2009). Vocabulary size and the Common European Framework of Reference for Languages. In B. Richards, H. Daller, D. Malvern, P. Meara, J. Milton & J. Treffers-Daller (Eds.), Vocabulary studies in first and second language acquisition (pp. 194–211). Basingstoke: Palgrave Macmillan.
6. Nation, I. S. P. (2006). How large a vocabulary is needed for reading and listening? Canadian Modern Language Review, 63(1), 59–82. doi:10.3138/cmlr.63.1.59