PTE Describe Image: how it is scored and how to answer

In PTE Academic Describe Image you get 25 seconds to study an image and 40 seconds to describe it. Content is marked first, and Pearson's rules classify pre-prepared or memorised material as irrelevant, so use a frame of ideas, not a script. Below: the marking, a safe frame, a phrase bank and two model answers of our own.

Describe Image is a speaking task in the first part of PTE Academic. You see an image and you describe it aloud. Pearson's research paper lists a chart, graph, diagram or picture. Here is how Pearson sets it out:

You can use the erasable noteboard while you prepare. Pearson says:

This page is about PTE Academic. Pearson describes it like this:

If you are taking PTE Core for Canada, read our PTE Core Describe Image page instead, because the number of questions and the scoring scale differ.

Content is the gate

In plain words: if the content is judged zero, nothing else is marked. Pronunciation and fluency only count once there is real content to score.

What good content looks like

Pearson's Score Guide scores content from 0 to 6. The wording is about what a listener could do with your answer:

The same guide describes the top level as a response that “describes the image fully and accurately and expands on the relationships between features of the image to provide a nuanced interpretation”, and the lowest levels as “a list of points identified on the image, without description or elaboration”. Our reading: describing is not listing. Say how the parts relate to each other.

Pronunciation and fluency

• The best responses are spoken at a constant and natural rate of speech with appropriate phrasing.
• Hesitations, repetitions and false starts will negatively affect your score.

Who does the marking

• Note: Pronunciation and Oral Fluency are AI-scored only.
• Note: Content is scored by both AI and human. If there is a disagreement between the AI and human, a second human will make the final judgment.
• Content, Pronunciation and Oral Fluency are all AI-scored traits. A human expert will also review Content before your final score for the task is confirmed.

Pearson's Score Guide also says this about the task:

• Some tasks such as Describe Image and Retell Lecture always require a human review of Content before scores are finalised.

A lot of people search for a Describe Image template. Before you use one, read what Pearson says about prepared answers. Pearson has said it in three places: the Score Guide, the format page and a research paper.

You saw above that the Score Guide classifies pre-prepared or memorised material as irrelevant, which makes the content score zero. The format page says it again from the other side:

The research paper is called Test integrity: Ensuring trust by integrating human judgment and AI systems. It is about PTE Academic, and it deals with Describe Image templates directly. It says:

• Any template that relies on long passages of memorized pre-scripted material is considered gaming.

The paper shows what it means by a memorised Describe Image template. The caption reads:

The example starts with a line like this one, which has nothing to do with any particular image:

Pearson's comment on that kind of template:

• This type of repetitive template is inappropriate because it relies heavily on memorized material, but it is also worth noting that the resulting response is not particularly sophisticated or high scoring.

Pearson also says it is not always a clear yes or no:

And it describes a human layer it added to its detection of memorised templates, which covers Write Essay, Describe Image and Lecture Retell (the paper's name for Retell Lecture). The paper says this came in 2024:

Our reading: Pearson does not say that planning your answer is a problem. It says long passages of memorised, pre-scripted material are. Pearson's Academic test tips carry a short note to the same effect:

• Your response should always be one you created.
• on test day, in your own words and describe information that is relevant to the image.

Frame or script? Our test

A quick test: if you could swap in a completely different image and your answer would still be mostly the same words, it is a script. If almost every content word changes and only the sentence shapes stay, it is a frame.

A note on dates. People look for a template with a year on it, but we found no template on any Pearson page we read. The Describe Image rules above are from the Score Guide, whose copyright line reads:

Pearson says the guide, together with its website, is the only official information about PTE Academic:

This is our frame. It follows the order Pearson's own tip recommends. Pearson says a well-organised answer is more likely to cover the main information, the details, and the implications:

Pearson's Academic test tips add one more reason to organise:

• By organizing what you say, it will be easier to tell the story of the image.

Six to eight sentences fit into forty seconds at a natural pace. Do not try to read every number. Pick the numbers that support the main point.

Everything in this table is ours. Pearson's format page says only that "An image appears on the screen", so the table covers the types most practice material uses. Use it to build a habit of choosing words, not to memorise lines.

Both examples are ours. We cannot show images here, so each image is written out as the data it would contain. The data is invented. The answers show structure and tone. We make no claim about what score either would earn.

Example one: a bar chart

The image. A bar chart titled "Adults who use the internet, by age group, 2024" for a made-up town.

Model answer (101 words). This bar chart shows the percentage of adults in a town who use the internet, by age group, in 2024. Overall, internet use falls as people get older. Almost everyone aged 18 to 29, 96 percent, goes online, and 93 percent of those aged 30 to 44 do too. After that the figure falls faster, to 85 percent for ages 45 to 59 and 62 percent for 60 to 74. The lowest figure, just 31 percent, is for people aged 75 and over. The gap between the youngest and oldest groups is large, which suggests older residents may need more support.

• Sentence one names the type of image, the topic, the group and the year, so a listener knows what they are hearing about.
• Sentence two is the main point, stated before any detail.
• The details are grouped, not read out one by one: the two highest figures together, then the next two, then the lowest.
• The figures are compared ("falls faster", "the lowest").
• The last sentence gives an implication, which is what the marking guide asks for.

Example two: a process diagram

The image. A diagram titled "How tap water is made safe" with six boxes joined by arrows: reservoir, screen, sand filter, chlorine added, storage tank, pipes to homes.

Model answer (95 words). This diagram shows how water is made safe to drink, from the reservoir to the home. There are six stages. First, water is collected in a reservoir. Next, it passes through a screen that removes large objects such as leaves. The water is then filtered through sand, and chlorine is added to kill bacteria. After that, the clean water is held in a storage tank, and finally it flows through pipes to homes. Each stage depends on the one before it, so the process turns raw water into a safe supply that reaches every household.

• It starts by saying what the diagram shows and how many stages there are.
• Sequence words (First, Next, then, After that, finally) show the order without long pauses.
• It adds a short fact for two stages where the diagram suggests it (leaves, bacteria), which makes it a description and not a list of box labels.
• The last sentence gives a short conclusion about the whole process.

1. Collect images. Cut or save a mix of charts, tables, maps and diagrams from news sites, reports and textbooks. Aim for variety. Twenty different images teach more than five images done four times.
2. Time yourself properly. Set a timer for the study time, then one for the speaking time. Do not pause the clock.
3. Record the answer. Play it back and count the hesitations, repeats and long pauses, because fluency is marked by a machine that listens for exactly those.
4. Run the listener test. Give the recording to someone who has not seen the image and ask them to sketch it. Pearson's top content level is a response from which a listener "could build a complete mental picture". If the sketch is wrong, find which sentence misled them.
5. Compare with a fresh answer. Describe the same image again a week later without notes. If the second answer sounds like the first, you were reciting.
6. Review the mistakes in a log. Note what you missed: the units, the trend, the conclusion.

A one-week routine (ours)

Official practice

Pearson's AI practice product says it lets you “Practice every question type across all four skills”. It is “Free to start, no payment required” and includes “2 diagnostic tests” in the free version, and its paid plans list “1000+ official questions”. Pearson's Scored Practice Test is a “2-hour practice test that feels just like the real test” at “Selling price: $35.99 USD”. Our PTE classes and courses page lists every option and what it costs.

• Reciting a memorised paragraph. It can be judged irrelevant and score nothing.
• Listing numbers. A string of figures with no comparison is a list of points, which sits at the bottom of the scale.
• No conclusion. Marking guidance asks for implications. One closing sentence is enough.
• Long pauses and false starts. A short, steady answer beats a longer one with gaps. If you slip, carry on.
• Starting before the microphone opens. Nothing you say before the tone is recorded.
• Stopping early. Use the forty seconds. If you run out of details, compare two you have already mentioned.

This page is not immigration advice. Pearson's Score Guide and format pages decide how the test is marked, and your visa or professional body decides which score you need. See our PTE score chart.

Common questions