The intermediate plateau is the stage, usually around B1, when you keep studying but stop noticing progress. Part of it is arithmetic: each extra thousand words covers less of a real text than the last. Breaking through usually takes more reading than before, deliberate review of less common words, and regular practice producing the language yourself.
Why does progress seem to stop at intermediate level?
Teachers and researchers describe a familiar pattern. A 2024 study by Utegulova and colleagues sums up the literature: Yi notes that at the start everything is new, which keeps motivation high, while Jack Richards observes that by intermediate level improvement slows, effort seems to vanish without trace, and learners grow frustrated. Richards lists the problems that typically show up at this stage:
- leaning on the vocabulary you learned first, and overusing it
- fluency that grows at the expense of complexity
- inadequate speaking and writing skills
- fossilised errors that persist despite progress elsewhere
- limited awareness of collocations, the words that naturally go together
The teachers interviewed in that study described the vocabulary side vividly: learners understand words when they read them but can't produce them when speaking, so they fall back on elementary words. Richards also gives a grammar example. A learner who is comfortable with the present and past has to reorganise that system to take in the perfect, and while that happens they tend to overuse what they already know.
The good news is in the definition. Murphy, cited in the same study, distinguishes the plateau, a temporary slowing of learning, from fossilisation, a permanent stabilisation. A plateau is a stage, not a ceiling.
Is the plateau real, or just maths?
Much of it is maths. Paul Nation measured how much of a typical English text each band of 1,000 word families covers:
| Word families | Share of written text covered | Share of spoken text covered |
|---|---|---|
| 1st 1,000 | 78–81% | 81–84% |
| 2nd 1,000 | 8–9% | 5–6% |
| 3rd 1,000 | 3–5% | 2–3% |
| 4th and 5th 1,000 together | 3% | 1.5–3% |
| 6th to 9th 1,000 together | 2% | 0.75–1% |
Your first thousand words change what you can understand almost overnight. Your fifth thousand might add one or two words per hundred you read. You are learning just as much, but the payoff on the page is smaller, so it feels like standing still.
The step sizes grow too. Studies reported by James Milton put typical learners of English at 1,500–2,500 of the 5,000 most frequent words at A2, 2,750–3,250 at B1 and 3,250–3,750 at B2. Contos's own thresholds follow the same shape: 500 known words to reach A2, another 1,000 to reach B1, then another 2,500 to reach B2. The longest climb on Contos is the one that starts at B1.
What changes in the text between A2 and B2?
We measured chapters 1–2 of every Spanish series on Contos at each level. The jump from A2 to B1 is the largest on almost every measure:
| Level | Words per sentence | Different word forms per chapter | Forms that appear only once in the level's free chapters | Forms needed to cover 80% of the text |
|---|---|---|---|---|
| A1 | 5.5 | 100 | 48% | 229 |
| A2 | 6.5 | 143 | 53% | 373 |
| B1 | 11.7 | 186 | 59% | 546 |
| B2 | 12.5 | 219 | 61% | 647 |
Sentences nearly double in length, and more of the vocabulary turns up only once. B1 text also asks more of your grammar, as this line from a B1 mystery shows, with a pluperfect, a preterite and two imperfects in one sentence:
Había visto su cara cuando la luz volvió a encenderse sola; no era sorpresa, era miedo.
That is the plateau seen from the page: more words you rarely meet, in longer sentences, built from grammar you understand in a textbook but haven't yet absorbed.
What actually helps you move past it?
- Read more than you did as a beginner. Words beyond the first few thousand turn up rarely, so you only meet them often enough in volume. Our extensive reading page has weekly targets.
- Review the rarer words on purpose. In Waring and Takaki's study, three months after reading a graded reader, learners could translate only 6% of the words they had met in it 15 to 18 times. Save the words that matter as flashcards and let spaced repetition bring them back.
- Read slightly above your comfort zone some of the time. Text that is all familiar consolidates what you know but adds little. The i+1 idea is a good guide to how far to stretch.
- Produce the language. Merrill Swain's output hypothesis holds that trying to speak or write shows you the gap between what you want to say and what you can say. Krashen disputes how much output adds, arguing that more input is the better strategy, but Richards's list of plateau problems is mostly about production, and reading alone won't fix it.
- Notice how words go together. When you meet a phrase that sounds natural, save the phrase, not just the word, so collocations become part of what you review.
Open a B1 series and read one chapter without stopping, then reread it with taps.
Read free in your browserFrequently asked questions
Is the intermediate plateau permanent?
No. Murphy distinguishes a plateau, which he calls a temporary slowing or stopping of learning, from fossilisation, where particular errors become permanent. The researchers in that study treat it as a temporary state that can be overcome with the right approach.
At what level does the plateau usually hit?
Researchers describe it as the move from intermediate to more advanced levels, which is why it is called the intermediate plateau. In CEFR terms, intermediate means the B levels, B1 and B2.
How long does it take to get from B1 to B2?
Cambridge English's guideline for learners of English is 350–400 guided learning hours to reach B1 from scratch and 500–600 to reach B2, so roughly 150–200 more hours. On Contos, B2 needs 4,000 known words against 1,500 for B1.
Should I switch to native books and TV?
Eventually, yes. Nation estimates 8,000–9,000 word families are needed to read a novel with 98% of words familiar, so native material is hard work at B1. Text written for learners at B1 and B2 bridges the gap. See how many words you need.
Does speaking practice really help with the plateau?
It's debated. Swain's output hypothesis says producing language makes you notice what you can't yet say. Krashen argues more comprehensible input is the better investment. Since typical plateau problems include weak speaking and overused basic vocabulary, most learners benefit from some of both.