How CEFR Levels Are Evaluated in Speaking and Writing

It is natural to assume a CEFR level is decided by how many words you know. It is not. Vocabulary matters, but tests that report CEFR-linked results score several things at once, and a wide vocabulary with weak control of the rest can still lead to a lower result than you expected. This guide looks at how that works, using public documents from two exam providers.

From a test score to a level

The CEFR itself is not a test, and it produces no scores. Exams are built to match its levels, and each exam turns performance into its own scale. Cambridge English, for instance, reports results on the Cambridge English Scale. Its exam pages list the ranges that correspond to levels: A2 Key at 120 to 139, B1 Preliminary at 140 to 159, B2 First at 160 to 179, C1 Advanced at 180 to 199, and C2 Proficiency at 200 to 230. IELTS reports band scores from 1 to 9. TOEFL iBT introduced a 1 to 6 band linked to the CEFR in January 2026. PTE Academic reports an overall score from 10 to 90, graded against the Global Scale of English.

None of these scales are the same, and the relationship between them is an estimate. That is one reason an organization that asks for "B2" usually names the exams it accepts. The framework itself is explained in what the CEFR is.

Speaking and writing are marked against criteria

For the productive skills, speaking and writing, a trained examiner or an automated system compares your performance to written criteria. IELTS publishes its band descriptors, which makes it a good example of how the criteria look.

For speaking, IELTS uses four criteria: fluency and coherence, lexical resource, grammatical range and accuracy, and pronunciation. For writing, it also uses four: task achievement (Task 1) or task response (Task 2), coherence and cohesion, lexical resource, and grammatical range and accuracy. The criteria are scored separately and contribute equally to the result for each skill.

Notice that vocabulary appears as one criterion out of four. Strong vocabulary helps in the others too, because it is hard to be fluent without the words you need, but the descriptors reward more than word knowledge.

What "lexical resource" actually looks at

Lexical resource is IELTS's name for vocabulary. The speaking descriptors show how it changes as the band rises.

At band 4, the descriptor mentions frequent inappropriate word choices and notes that the speaker rarely attempts to paraphrase. At band 5, the vocabulary is enough for familiar topics but flexibility is limited, and paraphrase is attempted without always succeeding. At band 6, the speaker has enough vocabulary to discuss topics at length, vocabulary may sometimes be inappropriate but the meaning stays clear, and paraphrase is generally successful. At band 7, there is some ability to use less common and idiomatic items, with awareness of style and collocation, though inappropriacies still occur. By band 8 and 9, the vocabulary is wide, flexible and precise, and used appropriately across topics.

Two ideas run through all of that. Paraphrase is the ability to say something another way when the exact word will not come. It is a vocabulary skill, because a larger vocabulary gives you more ways to go around a gap. Collocation is the habit of words to go together, as in make a decision and not do a decision, and the descriptors treat natural collocation as a sign of higher vocabulary skill.

Range, accuracy and fluency are different skills

Looking across the criteria in IELTS and in other CEFR-linked tests, three ideas keep appearing.

Range asks how wide a variety of words and structures you can draw on, not just how many words you know in isolation. A learner who uses the same safe words for everything has a narrower range than one who can vary them.

Accuracy asks whether the words and grammar you use are correct, including whether a word is used with its right shade of meaning. The speaking descriptors refer to errors in word choice and to how often errors get in the way of understanding.

Coherence and fluency ask whether ideas connect logically and whether you can keep going at a natural pace, recovering from a stumble without freezing. For speaking, the descriptors compare hesitation used to plan what comes next with hesitation used to search for words or grammar.

Why this changes how you study

Memorizing a long word list optimizes for range, and only range. It does little for accuracy, which means using a word correctly under real conditions, or for fluency, which means retrieving it fast enough that it does not interrupt your flow. That gap is why passively recognizing a word consistently overstates how ready you are to use it in an exam or a meeting.

What helps is practice that looks like the real task. Retrieval, trying to bring a word to mind before you check it, builds the kind of fast, accurate access the criteria reward. Research on test-enhanced learning, for example, has found that testing yourself leads to better long-term retention than rereading. Active recall versus rereading explains the idea in more detail. Using new words in your own sentences, aloud and in writing, covers accuracy and collocation as well.

What you can check in your own practice

You do not need an examiner to find out where your gaps are. Record yourself answering a question for a minute, or write a short paragraph, and then ask four questions of it. Did I use more than one way to say the main ideas, or did I repeat the same words? Were my word pairs natural, such as make a decision and heavy traffic? Did I keep going when a word was missing, by saying it another way? Did my ideas follow one another clearly? Even a rough answer to those questions points to the criterion you most need to work on, and the answer is often not "learn more words".

A note about automated scoring

Some tests, such as PTE Academic, use automated systems to mark responses, and say some responses are also reviewed by a human expert. The principle is similar: the system compares what you produce to scoring criteria. In that setting, clear, correctly pronounced, accurately used words tend to serve you better than ambitious sentences that wobble halfway through. The IELTS vocabulary guide looks at how this plays out for one specific exam.