Dare we say it, but the 2025 Reading SATs paper marks a welcome shift from the challenges of the past two years. The paper was still long, so children who struggle with stamina may have found it tough. However, from our perspective, it felt more straightforward than in previous years. After the notoriously difficult 2023 paper (has ‘it’ hit anyone yet?) and the still-demanding 2024 version (Texas trauma, anyone?), this year’s paper felt more balanced, and… boring? Did we hear that right?
Could it be that children were presented with a paper that didn’t leave masses of them in tears and included more accessible texts? Could the reading gods finally have been in our favour after two tough years?
Well….no. We’ve read the mark scheme.
While the paper itself may have been more accessible, the mark scheme is a different story. We had one wish: clarity and consistency. Sadly, that’s not what we got. Instead, the mark scheme is fraught with ambiguity. It’s not all doom and gloom though. More on that shortly…
We’ll now give a brief overview of the entire paper, look at each extract, and share some stats to evidence our impressions. See what we did there?!
Progressively Harder Extracts
The three texts gradually increased in complexity, offering less confident readers a chance for early success. The texts included: a biography with clear non-fiction signposting; a Science Fiction/Mystery narrative, with dialogue to advance the action; and a historical fiction extract taken from Longbow Girl by Linda Davies, of which we may have heard rumblings about its appearance in a certain Accelerated Reader programme…
This gradual increase in complexity is a welcome return to the format of SATs pre-COVID, which seemed to be lost over the last few years when harder extracts and questions appeared too early, leaving children demoralised. Compared to 2023’s dense vocabulary and overwhelming content, we believe that the 2025 paper was more readable and better suited to a broad Year 6 audience.
That said, it’s hard to ignore how often some children described the extracts as ‘boring.’ While more accessible, the extracts still lack the excitement and engagement found in modern children’s literature. With so many outstanding children’s authors available, SATs extracts still feel dated and disconnected from what truly inspires young readers.
Inference Heavy
The 2025 reading paper continues to follow the familiar pattern of placing most emphasis on Inference and Retrieval, which together account for 39 out of 50 marks. Inference is the most heavily weighted domain, with 24 marks (the highest it has been in the last three years) showing a strong focus on pupils’ ability to interpret and read between the lines. Retrieval is still a key element, though its weighting has dropped slightly to 15 marks, compared to 19 marks in 2024 and 16 marks in 2023. Word meaning questions remain important, contributing 6 marks. This is a slight increase from5 marks in 2024, though still lower than the 9 marks seen in 2023.
The other content domains receive much less attention, with no marks given for prediction or meaning as a whole. Summarising appears every year but carries very little weight with just 1 mark in 2023 and 2024, increasing slightly to 3 marks in 2025. Comparison questions were not included in 2023, peaked at 3 marks in 2024, then dropped to 1 mark in 2025. The effect of language domain was only assessed in 2025, with 1 mark, and was absent in the previous two years. Overall, the differences across the three years are quite small, with the 2025 paper providing a strong emphasis on inference as the key comprehension skill, while giving less weight to other domains.
Total Word Count = 3,084 words
The 2025 paper had a similar word count to 2024 and was shorter than 2023. However, it somehow felt like there was less to read than last year — likely because the vocabulary was simpler, and the extracts were more straightforward.
With this word count, children had around 35 minutes to read all the extracts and questions (based on a reading speed of 90 words per minute). That left approximately 25 minutes to answer 40 questions — giving them about 37.5 seconds per question. Since the readability of the extracts got progressively harder, fluent readers may not have found this pace too daunting.
Extract One: A Life-Changing Game Genre: Biography
This extract tells the true story of Phiona Mutesi, a Ugandan chess player whose life inspired the book The Queen of Katwe by Tim Crothers and the 2016 Disney film of the same name. The non-fiction text was well-structured with clear subheadings and focused mainly on retrieval questions. These questions guided readers to the right sections to find their answers.
At first, this extract seemed fair and accessible. However, after reading the mark scheme, we found it to be very ambiguous (actually, the most ambiguous of all three extracts). The accepted answers often differ by just one word, making it feel like a tricky game of ‘spot the difference.’ It’s disappointing that such strict marking appears on a paper that otherwise seemed accessible for less fluent readers.
Mark allocation for this extract:
Retrieval (2b): 5 marks
Inference (2d): 5 marks (questions 3, 4, 7, 8)
Explain meaning of words in context (2a): 4 marks (questions 2, 6, 9, 12)
Examples of marking ambiguity
These examples highlight some of the ambiguity in the mark scheme. Our focus here is on the unclear parts, but don’t lose hope: many of these questions offer other correct options for children to earn marks.
Question 3: How can you tell that people from Uganda were not familiar with chess?
Accepted: “There was no word for it.”
Not accepted: “They hadn’t heard of it.”
This was a retrieval question, so children probably got it right if they copied it down directly. However, if children used inference and said “they hadn’t heard of it,” they’d be marked wrong, even though it feels like a reasonable answer.
Question 5: What is it important to have when learning something new?
Accepted: “Someone to help you learn.”
Not accepted: “Someone to help you.”
Even though the question itself refers to learning something new, children still have to include the word learn in this particular answer (other answers are accepted). Just saying “someone to help you” isn’t enough — even though the context makes the meaning clear. You’d think the wording of the question would allow for some flexibility, but not in this case.
Question 7: Why did Robert start teaching chess?
Accepted: “Because he realised that the children had no desire to play football.”
Not accepted: “No children had the desire to play football.”
Accepted: “Because kids lost interest in football so he taught chess.”
Not accepted: “Because no one liked football so he taught chess.”
We don’t really know what to say about this one other than…Good luck, markers.
Question 8: Why did Phiona play carelessly when she started playing chess?
Accepted: “She’s so eager.” “She’s too eager.”
Not accepted: “She’s eager to win.”
It looks like only answers that intensify eager with adverbs like so or too are accepted. Why? Because showing that she was too eager is the reason for the carelessness. Arguably, so does saying ‘she’s eager to win’…but what do we know!
Overall, the paper may appear accessible at first glance, but the mark scheme has provided some strict and often unclear rules. While we don’t want to sound overly negative, the level of ambiguity is frustrating — especially given that this was a more accessible paper for less fluent readers. It’s a shame that such inconsistencies appear in a mark scheme that accompanies one of the more inclusive texts.
Extract 2: In the Cave Genre: Science fiction/Mystery narrative
This mystery-driven extract required readers to infer meaning, particularly about the strange machine the characters discover. Language like hovered, inaudible, and curiously added moderate difficulty. There was also a straightforward 3-mark question asking for impressions and evidence of the different personalities, using wording that has been used many times in previous SATs papers. However, this paper had the highest number of inference questions, many of which were similarly worded. This could have confused some children, especially with repeated references to the buttons and lights.
Mark allocation for this extract:
Inference: 8 marks
Retrieval: 5 marks
Summarising: 3 marks
Meaning of words in context: 1 mark
How meaning is enhanced by word choice: 1 mark
Examples of tricky questions
Q22: Why did Tom shrug?
Accepted: “He doesn’t know what the machine is” or “he thinks it’s obvious what the machine is.”
Children might think it means “he doesn’t care” or “he’s not bothered,” especially compared to Geoff, who seems more enthusiastic. However, to answer correctly, children need to show a full understanding of Tom’s personality and that he’s more cautious.
Q24: How did Geoff work out that the lights were buttons?
Accepted: “he ran his finger over them” or “he stared intently at them,”
Children can’t just say “because he pressed one”— that’s too simple (!) The question asks how he worked it out.
Q26: Impressions of Tom and Geoff’s personalities
Tom: anxious, sensible, observant, curious
Geoff: adventurous, reckless, curious, excitable, stubborn, observant
This question should be straightforward if children focus on describing personality traits. Geoff has more traits mentioned, and both boys can be described as curious. Tom’s personality might be trickier to understand if children think, like in the shrugging question, that Tom is simply unbothered, they could lose marks.
Extract 3: Longbow Girl Genre: Historical fiction
This extract, from Longbow Girl by Linda Davies, caused quite the stir with rumblings of it having also been included in the accelerated reader programme. Some of your children may have read it before…whether this is helpful or not is open to interpretation! Arguably, this was the most difficult text with phrases like cool focus flooded her veins and unbearable tension showing ambitious vocabulary and complex sentence structures. The use of inner monologue also added to the understanding of the character’s emotion and the suspense. This was an extract that could potentially challenge some of your readers.
Helpful marking…(yes, you’ve read this heading correctly!)
A few helpful features stood out in mark scheme for this extract. Thankfully, the words “left” and “right” were included in the text, so children didn’t need to know them beforehand to answer that question. Also, for the first time, the mark scheme included a model answer — specifically, an example of why Merry was considered experienced, with supporting evidence. This was a welcome and useful change. Separate marks were also given for different aspects of Merry’s knowledge (like past learning, techniques, and rules), rather than grouping them all together. In the past, teachers across the country may have shouted, “Those answers are too similar; they won’t get a mark!” But here we are in 2025: it might be time to rethink everything we thought we knew about mark schemes…
Mark allocation for this extract:
Retrieval: 5 marks
Inference: 11 marks
Making comparisons within the text: 1 mark
Examples of questions and mark scheme
Q32: Why was the crowd stunned?
Accepted answer: “Because she did good.”
This may be grammatically jarring, but since the answers in italics are based on children’s trial answers, it reflects what markers are likely to encounter.
Q35: Why were the rules changed after the first round?
Accepted: “To make it more interesting to watch.”
Not accepted: “To make it interesting to watch.”
The difference lies in the adverb of degree “more,” which shows that the change was made to increase interest from round one to round two. Once again, as in the first extract, a single word seems to be the difference between getting a mark or not. At least this time, it makes a little more sense to us.
Final Thoughts
Overall, this year’s paper showed a genuine effort to be fair and more accessible, giving students a realistic chance of success. While the high word count still demands stamina, the structure and question design suggest that examiners may be responding to past feedback. However, the mark scheme’s persistent ambiguity detracts from this progress, leaving a sense of frustration. Fortunately, much of this uncertainty is limited to the first extract and eases as the paper progresses, with clarity notably improving — particularly in the three-mark questions. The rise in inference-style questions highlights the growing emphasis on this reading domain: to succeed, children must have strong fluency to enable them to gain a deep understanding of the texts; they cannot rely on retrieval skills alone.
To sum up:
Accessible for most.
Inference-heavy.
Marking? It’s a shift, and one that might make teachers question everything they thought they knew about mark schemes.
Welcome to 2025, fellow educators. Now, what’s your take?
If you would like to download the Reading Sats Papers you can do so here on Literacy Shed Plus