KS2 Reading SATs papers 2026: analysis from Sophie Bartlett

Sophie Bartlett is back with her KS2 SATs Reading analysis for 2026. She delves into key data trends from this year’s paper, including: the fiction vs non-fiction text split; how this year’s word count compares to previous years; inference overload and unnecessarily ambiguous question wording. 

If you’re after a complete guide to SATs, including when results are released, check out the Ultimate guide to KS2 SATs. 

Share this analysis with your team...

Sign up to our mailing list and get a PowerPoint slides version straight to your inbox.

Get the slides

Well OWL be damned… there goes another year where the nation’s year 6s were spared (or perhaps, unfairly denied?) the pleasure of poetry analysis. Which begs the question, which will be found first: a poem in the SATs, or Earhart’s plane? My money’s on the latter (unless you believe the coconut crab theory… now that’s a story for another time!)

Key takeaways: KS2 SATs Reading 2026

• This year was the first time the opening two texts of the KS2 SATs reading paper have been fiction. 
• Reading comprehension alone is not enough. Children increasingly need familiarity with the structure and language of SATs questions themselves.
• Overall, teachers felt the texts were fair and engaging, but the questions were unusually inference-heavy. 

Reading paper 2026: what was the word count?

The 2026 SATs reading paper had a word count of 1,961. My genuine (but less entertaining) theory of why no poem is that it’s simply a struggle to meet the required word count… The English reading test framework states that the selection of texts should comprise of 1,500 - 2,300 words. So, if we think back to the word count of the 2018 paper (the only test so far with a poem) it barely scraped 1,500 words. This makes sense when you consider that the poem (its second text) was about 400 words shorter than the average second text.

Average word counts for each text (based on data from 2017)

• First text: ~580 words
• Second text: ~645 words (discounting the 2018 poem)
• Third text: ~710 words

The texts are ordered by increasing reading demand, and so it was a surprise to see a non-fiction piece as the final text this year - in fact, this is the first time it’s happened!

Word counts for SATs reading paper texts from 2017 to 2026

Key: Fiction Non-fiction Poetry

Whilst the paper may have felt wordier to some, the texts themselves totalled around 2,000 words - roughly the midpoint of previous papers, which have ranged from approximately 1,500 words in 2022 to 2,200 in 2019. Even when the question booklet is included, the total comes to around 3,000 words; this is still lower than papers such as 2019 and 2023, which were closer to 3,200.

Word counts for SATs reading papers (questions and texts inclusive)

Reading SATs: fiction vs non-fiction

Generally, comprehension questions based on non-fiction texts seem less demanding than those based on fiction. There’s less room for nuance: more, 'What year did this event happen?' and less, 'What deep emotional state was the author hinting at through the character’s weird little hand movement halfway through paragraph three?'

It’s also the first time the opening two texts have both been fiction. For pupils who struggle to finish the test in time, these are usually the sections most likely to secure the marks needed for EXS: together, they’ve historically been worth 31-34 marks, while the ‘pass’ mark has sat between 24-29. As mentioned earlier, non-fiction questions generally tend to feel more straightforward, making those marks easier to secure. This year, having two fiction texts back-to-back may have slowed children down - an opinion also reflected in the teacher feedback.

Want the 2026 maths insights?

Check out Sophie's analysis of the 2026 maths papers.

Read more

Teacher response after the Reading paper 

How did teachers find the 2026 reading paper compared to last year? Over 1,000 teachers voted…

How did teachers think their pupils found the 2026 reading paper? Over 1,000 teachers voted…

Overall, the teachers I asked generally felt the texts themselves were fair and engaging, but that the questions were unusually inference-heavy, with fewer straightforward retrieval questions than in previous years. Many commented that the paper was time-consuming and writing-heavy, with several pupils - particularly slower or borderline readers - struggling to finish within the allotted time. Teachers also said:

• 'Loved the texts - thought they were great for our age range'
• 'Why does everything you say go out the window in a test?'
• 'So much to read! Really heavy inference. Really slowed kids down'
• 'Texts good. Had to work hard for some answers. A lot of mine struggled with timings - v slow!'
• 'A lot of text. A lot of inference. A lot of tears from chn who find it slow to read/answer questions'
• 'Texts were great but questions poorly worded and so much inference'

As for the questions feeling 'inference-heavy', or 'retrieval-light', well… 

Analysis of reading content domains

The prevalence of retrieval (2b) and inference (2d) questions continues, although (perhaps surprisingly given initial teacher perceptions (see below) before the mark scheme release) inference is at its lowest proportion since 2019. Vocabulary (2a) questions remain lower than in previous years (peaking at 20% in both 2017 and 2018), although there is naturally overlap between vocabulary and inference skills, regardless of how questions are classified.

This year also saw the highest proportion of whole-text (2f) questions on record, with 2 marks allocated to this domain (and both in the second text of the paper). As in previous years, there were no prediction (2e) questions; this strand has only ever accounted for 1 mark overall, back in 2022.

Let’s face it, whether a question is labelled as ‘retrieval’, ‘inference’ or ‘vocabulary’, its level of difficulty is completely subjective based on the child - their ability to decode; their background knowledge of the text; their emotions that day; whether or not Susan, the Random Invigilator they’ve never seen before, is hovering over their shoulder for the third time within the last 30 minutes…

I couldn’t (and would be petrified to) imagine what a 10-year-old was thinking as they sat this test, but this summarises my thought process as I read through the paper:

'An animal! Could’ve guessed. At least it isn’t an obscure one they’ll never have heard of….

…Ooh, Matilda! They’ll definitely know… oh wait, not that Matilda. Still a young, clever girl called Matilda - just not the one they’ll definitely all be thinking of. Cool…

…Gosh, it really does take an insane level of talent to make Amelia Earhart’s story read like the terms and conditions section of a website…

…Hold on, where were all the ‘Find and copy’ questions?!'

On a more positive note, did anyone else also appreciate the aesthetics of each text fitting nicely on a double-page spread? Aside from being very visually satisfying, I do think this makes it easier for the children to answer the questions. Few things are more frustrating in a test than constantly having to flip pages back and forth. 

So, how did each individual text hold up under closer scrutiny?

Breakdown of fiction text 1, Owl in Danger

The first text, worth 16 marks and around 570 words (similar to opening texts from previous SATs papers), was an extract from Owl in the Office (Animal Ark) by Lucy Daniels (1995). I thought it was a friendly text to open with - not visually overwhelming, short paragraphs and some helpful pictures. The questions were also fair, although there were noticeably more questions encouraging children to consider the ‘whole text’ - I would’ve thought these would relate to the content domain 2f, but both these questions below are labelled as inference.

The mark scheme was fair in both cases, crediting a range of valid points and requiring pupils to identify only two correct answers.

Although the text itself was relatively straightforward, only 1 of the 16 marks awarded was for retrieval (often considered the more accessible question type). In contrast, 10 marks were allocated to inference, meaning this text alone accounted for almost half of all inference marks overall (21 of the total 50).

Want the 2026 GPS insights?

Check out Sophie's analysis of the 2026 GPS papers.

Read more

Breakdown of fiction text 2, Matilda’s Invention

The middle text, worth 16 marks and around 630 words (similar to middle texts from previous SATs papers), was an extract from The Matilda Effect by Ellie Irving (2017). On first impression, this looked like a friendly middle text: slightly word-heavier than the first (as expected), but with colourful pictures and interesting, age-appropriate content. The questions though… oof.

I tripped up on the very first question for this text.

Maybe it’s just me, but I read this as, ‘Why does Matilda choose the clothes she wears?’ (as opposed to her grandparents, or anyone else). Yes, I know, common sense tells you this isn’t the correct interpretation - especially as the first instruction is to ‘read the first paragraph’, where the grandparents aren’t mentioned, but this could just be a reference to the description of the clothes. A clearer wording of the question could have been, ‘Why does Matilda choose to wear those particular clothes?’ 

Just like 2025’s reading analysis, this paper required children to be secure in their knowledge of language; in this case, to know that ‘creativity’ could be a synonym for good imagination, and ‘materials’ for a pile of junk.

Here’s some more wording that caught me off guard.

This question seems fairly straightforward, right? Read the relevant section of text and think about what you would have written if you were sitting the test.

The mark scheme accepts… but also feels like a perfectly reasonable response… at least until you reach question 22.

The answer to this is obviously asking for a reference to the ‘Ask your teacher for details’ line, which makes you realise that question 21 probably isn’t asking for the same answer. But should we really expect children to apply that kind of test-writing logic under timed conditions?

This feels like another example where slightly clearer wording could have avoided unnecessary ambiguity. Something like, “What must pupils enter for the competition?” would have been much more precise.

There were two more ‘whole text’ questions in this section (categorised as 2f and 2d respectively), one of which was our first of two three-markers in the paper.

Although this text felt more challenging than the first, almost half of the marks were for retrieval (7 of 16 in this section), with only 5 marks allocated to inference. This highlights that, while retrieval questions are often considered more accessible, difficulty ultimately depends on context rather than question type alone. 

Thoughts on unnecessary ambiguity in SATs Reading paper questions

Why do I care so much about unclear wording in questions, you might ask? Well, I’m not one to catastrophise (although my husband would tell you otherwise), but, picture this scenario:

• A 10-year-old misinterprets the wording of a relatively ambiguous question.
• They lose the mark.
• That one mark is the difference between a ‘pass’ and a ‘not pass’ (I refuse to use the word ‘fail’ when it comes to SATs).
• The child is part of a cohort of 12, meaning they represent over 8% of the school’s data. Suddenly, 75% passing (9/12) becomes 67% passing (8/12), a drop equivalent to around eight children in a Year 6 cohort of 100.
• 67% sits below the national average, which tends to hover around the mid-70s. The published conversation then becomes 'below national' or 'lower than last year', rather than any meaningful discussion about context.
• The school is pressured to show how it plans to 'improve reading outcomes'. Changes are introduced, not necessarily because they will improve anything, but because schools are expected to be seen doing something. Workload increases, stress rises, and attention is diverted away from other areas desperately needing support, simply because they are harder to measure.
• As absurd as it sounds, this child’s result may even contribute to lower GCSE target-setting later on, thanks to the education system’s love affair with data

… all because of slightly shoddy question wording.

That might sound dramatic, but for many small schools, it’s a completely plausible scenario. And that’s why I care so much about individual SATs questions.

After the slightly brutal middle section, the final text was surprisingly kinder.

Breakdown of text 3, non fiction, Amelia Earhart

The final text, worth 18 marks and around 770 words (similar to final texts from previous SATs papers), was a non-chronological report about Amelia Earhart. This text was definitely the most visually overwhelming - a lot of words, chunky paragraphs, and features that famously engage children’s interest: monochrome pictures and a map.

Despite the more complex language (including vocabulary such as pioneering and unconventional), I felt the questions were relatively fair and straightforward - much more so than the previous text. In fact, almost half of all the retrieval questions in the paper came from this final text (8 out of the 18 marks allocated to this section).

I particularly liked this one-mark retrieval question - it took some close reading to come to the answer of ‘west to east’ (as opposed to ‘east to west’ on her previous flight)!

The test ended on the second of the two three-markers - children had five possible answers to choose from, and provide evidence for, here (that she was determined, brave, ambitious, organised or patient).

The evolution of the 3-mark question

It’s been interesting to see how the test writers have handled the infamous 3-mark questions over time. We’ve only ever had two per paper, despite the STA stating there could be up to four, and they’ve never appeared in the first text. As most of us now know, these hefty questions generally require two points, each supported with evidence, to secure all 3 marks.

2016 was the first SATs paper under the new curriculum (does the phrase ‘milled around in bewilderment’ still strike fear into the hearts of any other seasoned year 6 teachers, or is that just me?). The 3-mark questions were pretty intimidating: sometimes taking up almost an entire page and phrased along the lines of, ‘Explain your choice fully, using evidence from the text’, followed by a box containing nine ominously blank lines. Unsurprisingly, many children did not even attempt them.

In 2017, the tests at least started giving children a little more direction, adding wording such as, ‘Explain two ways, giving evidence from the text to support your answer’. However, the daunting wall of blank lines remained.

It was not until 2018 that they finally seemed to settle on a more helpful structure. This was the first year the now-familiar grid layout appeared, with headings such as ‘Impression’ and ‘Evidence’. It gave children a much clearer framework for organising their responses, and variations of this format have remained ever since, including in question 38 this year.

KS2 SATs Reading paper: results

The ‘pass mark’ has historically ranged from 48% to 58% - the widest margin of all the papers, whilst the threshold for the higher standard (technically, ‘GDS’ only applies to writing, but a scaled score of 110 is commonly used as the equivalent benchmark) has remained within a range of 76% to 82%.

As for this year, most of you predicted the mark for EXS would drop by 1. We’ll find out for sure in July!

What do you think the ‘pass’ mark for the 2026 reading paper will be?

Reading 2026: conclusion

Perhaps the biggest takeaway from this year’s paper is that reading comprehension alone is not enough. Children increasingly need familiarity with the structure and language of SATs questions themselves, particularly when it comes to ambiguous wording, ‘whole text’ questions and the multiple three-markers.

Overall, I do think this year’s reading paper was fairer than some we’ve seen before. The texts themselves were engaging and refreshingly age-appropriate, despite somehow managing to drain all suspense from one of the greatest disappearance stories of all time.

The bigger issue for me was not difficulty, but precision. When so much weight is attached to SATs outcomes, the wording of questions matters enormously. A single ambiguous phrase might seem insignificant in isolation, but in reality, those tiny moments can ripple far beyond one mark on one paper.

Perhaps that is the strangest thing about the KS2 SATs reading paper: we are supposedly assessing children’s reading ability, yet increasingly, success can depend on understanding the logic of the test itself.

Anyway. Same time next year, when we all once again convince ourselves the poem is finally coming - unless that, too, has fallen victim to the coconut crabs.