VSTEP · Guides · November 6, 2026 · 6 min read
VSTEP exam format: Listening, Reading, Writing and Speaking part by part (B1–C1)
VSTEP (Vietnamese Standardized Test of English Proficiency) is Vietnam’s national English test. Its full name is VSTEP.3-5, because one paper measures three levels of the country’s six-level foreign language framework (Khung năng lực ngoại ngữ 6 bậc dùng cho Việt Nam): bậc 3, 4 and 5, which match CEFR B1, B2 and C1. You do not register for a level. Everyone sits the same test, the tasks get harder as you go, and your overall score decides which level, if any, appears on the certificate.
The format was set by the Ministry of Education and Training in Decision 729/QĐ-BGDĐT of 11 March 2015, and that format still applies as of November 2026. The rules for organising the exam have changed since then (see registration and fees), but the parts, question counts and timing below have not.
| Skill | Parts | Questions / tasks | Time | Share of the overall score |
|---|---|---|---|---|
| Listening | 3 | 35 multiple-choice | about 40 min, incl. 5 min to transfer answers | 25% |
| Reading | 4 passages | 40 multiple-choice | 60 min, incl. transfer | 25% |
| Writing | 2 tasks | a letter/email and an essay | 60 min | 25% |
| Speaking | 3 parts | interaction, solution discussion, topic development | about 12 min | 25% |
Each skill is scored from 0 to 10, and the overall result is the average of the four. The details of the calculation are in VSTEP scores and levels.
Listening: 35 questions, each recording heard once
The single most important fact about VSTEP Listening is that every recording is played only once. Candidates who have prepared with IELTS or school tests, where audio is often repeated, are caught out by this more than by any vocabulary.
- Part 1 (Questions 1–8). Eight short announcements, instructions or messages: a station announcement, a voicemail, a teacher explaining an assignment. One question each.
- Part 2 (Questions 9–20). Three conversations, four questions each. Typical settings are students planning a project, a customer and a receptionist, or colleagues discussing a change at work.
- Part 3 (Questions 21–35). Three talks or lectures, five questions each, on academic but accessible subjects. Expect questions about the speaker’s opinion and about what is implied, not only facts.
All 35 questions have four options (A–D). The difficulty rises from Part 1 to Part 3, roughly from B1 towards C1. Our Listening guide has a worked example and a method for reading questions ahead.
Reading: four passages, 40 questions, 60 minutes
There are four passages with ten questions each. Together they come to about 1,900–2,050 words, so the later passages are longer as well as harder: the first is usually a practical, everyday text and the fourth reads like an academic article. Questions test detail, the main idea, the writer’s opinion or attitude, inference and the meaning of words in context, with four options each.
Sixty minutes for 40 questions leaves about 15 minutes per passage, including the time to transfer answers. In practice, strong candidates finish Passage 1 in 10–12 minutes and bank the extra time for Passage 4.
A vocabulary-in-context question looks like this (our own example):
Paragraph: “When the community garden first opened, few residents took an interest. Within two years, however, it had become the busiest public space in the district, and the council decided to replicate the idea in three other neighbourhoods.”
The word “replicate” in the paragraph is closest in meaning to: A. advertise B. repeat C. reject D. fund
The answer is B. The sentence says the council took the same idea to three other places, which means doing it again. “Fund” is the tempting distractor, because a council might well pay for gardens, but the paragraph says nothing about money; it is about copying a successful idea. That is the general rule for this question type: put each option back into the sentence and check it against the logic of the paragraph, not against what sounds plausible in real life.
Writing: two tasks, 60 minutes
| Task | What you write | Length | Suggested time | Weight in Writing |
|---|---|---|---|---|
| Task 1 | a letter or an email (formal or informal) replying to a situation | at least about 120 words | about 20 min | one third |
| Task 2 | an essay on a general topic | at least about 250 words | about 40 min | two thirds |
Both tasks are marked on four criteria: task fulfilment, organisation, vocabulary and grammar. Because Task 2 carries twice the weight, spending 30 minutes polishing the letter is one of the most expensive mistakes on the paper. Task 2 is covered in detail in our Writing Task 2 guide.
Speaking: three parts, about 12 minutes
| Part | Task | Usual timing (not an official rule) |
|---|---|---|
| 1. Social interaction | questions on two everyday topics | about 3 min |
| 2. Solution discussion | a situation with three possible solutions: choose the best and explain why the others are weaker | about 4 min, with 1 min to prepare |
| 3. Topic development | talk on a topic using a mind map of ideas (or your own), then answer follow-up questions | about 5 min, with 1 min to prepare |
The Decision does not publish a per-part timing; the figures above are what preparation materials and test centres commonly describe. The criteria usually cited are grammar, vocabulary, pronunciation, fluency, and discourse management or content.
Part 2 is where many candidates lose easy marks: they explain why they like one option and forget to say why the other two are worse. Here is how one minute of preparation can be used, with our own example task:
Situation: Your English club wants more members. Three ideas have been suggested: a stand at the university open day, a page on social media, or free trial lessons every Friday. Which is the best choice?
A workable plan, written as notes in 60 seconds:
- Choice: free trial lessons.
- Why it works: people experience the club, not just read about it; regular (every week), so friends can bring friends.
- Open day: only one day a year; busy, noisy, people collect leaflets and forget them.
- Social media: cheap, but our posts compete with thousands of others; likes do not turn into members.
- Close: start with trial lessons, use the page to announce them.
That plan gives a clear structure for two minutes of speech and covers all three options, which is exactly what the task asks for.
How one test gives three levels
Because the paper climbs from B1 to C1 inside every skill, a candidate at B1 will answer most of Listening Part 1 and Reading Passage 1 correctly and fewer questions later on, while a C1 candidate will also handle the lectures and the academic passage. The same idea applies to Writing and Speaking: examiners judge how far your performance goes, not whether you completed a level-specific task.
The thresholds, as of November 2026:
| Overall score | Result |
|---|---|
| below 4.0 | no level awarded |
| 4.0–5.5 | bậc 3 (B1) |
| 6.0–8.0 | bậc 4 (B2) |
| 8.5–10 | bậc 5 (C1) |
The practical consequence is that you should not skip the hard parts if you only need B1. Every correct answer in Part 3 or Passage 4 counts as much as one in Part 1, and a few of them can carry a borderline average over 4.0 or 6.0.
On the day
Some centres, including ULIS (VNU Hanoi), run VSTEP as a computer-based test, so practise typing your essay and reading passages on screen, not only on paper. The order of the skills, the arrival time and the documents to bring are set by each centre; they are in the notice for your session.
Where to start
Read this page once, then take a timed mock of one skill to find out where your marks go. Our VSTEP mock tests follow the format above and give an estimated 0–10 score and level; they are practice estimates, not official results. If you have several weeks before the exam, the preparation plan suggests how to divide them.
Nomto is an independent preparation platform and is not affiliated with the Ministry of Education and Training or any VSTEP test centre.
Also in Tiếng Việt