KS2 Maths SATs papers 2026: analysis from Sophie Bartlett

Gather round, because Sophie Bartlett is back with her much-loved analysis of the KS2 maths SATs papers for 2026. Jump into: data trends galore, which content domains dominated the paper and Sophie’s take on some of those controversial reasoning questions. 

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 wasn’t that an emotional rollercoaster? An arithmetic paper sent from the SATs gods, a reasoning paper 2 to make us weep and a paper 3 that, after such a gruelling few days, everyone was too tired to form a strong opinion on, really.

Across the three papers, there seemed to be a big emphasis on mathematical flexibility and efficient strategy choice. In several places, children who could recognise patterns, manipulate numbers mentally or adapt methods confidently were at a real advantage over those relying purely on rehearsed procedures. As always, there were questions I loved, questions I hated, and a few that seemed to divide opinion… 

Key takeaways: KS2 maths SATs 2026

• The 2026 papers rewarded mental maths and pattern recognition over rote procedures, especially in Arithmetic and Reasoning Paper 2.
• Paper 2 was the most hated. 54% of teachers found it too hard, citing multi-step 1-mark questions, complex wording, and children in tears.
• Key topics are still missing. Scale factors, reading scales, algebra missing numbers (A1), and circles haven't appeared in 3-5 years. Converting metric units was absent for the first time ever.
• Paper 3 was a relief. Teachers called it "fairer," "less wordy," and a "breeze" compared to Paper 2.

Maths SATs 2026: Analysis in numbers

A closer look at this year’s maths SATs papers reveals how questions were distributed across both year-group curricula and content domains, highlighting some notable shifts in emphasis compared with previous years. 

Maths SATs 2026 content weighting, by year group 

Year 6 consistently contributes the largest proportion of questions in the maths SATs paper, typically ranging from 37% to 47%. This year’s tests show an increase from last year, reaching a proportion similar to that seen in 2024. In contrast, year 3 content makes up the smallest share, typically between 7% and 12%. Whilst we might expect a steady increase in the proportion of questions from year 3 to year 6 - which has been the case this year - this is not always true: in both 2017 and 2024, there were more questions drawn from the year 4 curriculum than from year 5.

This year’s tests also contain more questions from the UKS2 curriculum than any previous papers (with the exception of 2018): 71% of the test content has been drawn from the year 5 and year 6 curricula.

Maths SATs 2026 question analysis, by content domain

As in all previous SATs papers, questions from the Calculations content domain dominate the test, closely followed by Fractions, Decimals and Percentages, with this year’s proportions exactly matching those seen last year. Both the Ratio and Proportion and Statistics content domains saw an increase in questions this year, each reaching levels not seen since before the two-year break in SATs papers during Covid.

(It’s worth noting that the KS2 maths test domains are weighted unevenly. Domains with more objectives, such as Calculations and FDP, receive the most marks because they contain the most content. However, this is not always consistent: Measurement has 44 objectives but has accounted for less than 10% of marks in recent years.)

SATs 2026 Maths Papers: content weighting

As in previous years, over half of the content in every KS2 maths paper has come from the years 3, 4 and 5 curricula. Based on the pass marks from most previous papers, children could theoretically meet the expected standard through secure knowledge of year 3-5 content alone. Similarly, in theory, pupils could also meet the expected standard through knowledge drawn almost exclusively from the Calculations and FDP domains. As long as the ‘pass’ mark remains below 58% - and historically it has never been higher than 55% - it is still theoretically possible for a child to achieve EXS without answering any year 6 content correctly, highlighting the importance of secure foundational maths knowledge across KS2.

Analysis of Paper 1: arithmetic

In every arithmetic paper so far, including this year’s, there has always been:

• 36 questions, 32 of which are worth 1 mark
• Two long multiplication questions, each worth 2 marks - one is 3-digits by 2-digits, one is 4-digits by 2-digits
• Two long division questions, both in the second half of the paper, each worth 2 marks - one is 3-digits by 2-digits, one is 4-digits by 2-digits

This year’s paper covered everything you’d expect in an arithmetic paper (some topics have been omitted in previous years, such as multiplying a mixed number from both 2022 and 2023). I genuinely think the 2026 arithmetic test has been my favourite yet - although, let’s face it, the competition isn’t exactly fierce…!

Want the 2026 Reading insights?

Check out Sophie's analysis of the 2026 Reading paper.

Read more

Arithmetic paper 2026: Question 2

We’d always expect to see addition and/or subtraction in the first double spread of questions; they couldn’t quite help themselves with this one though, could they. God forbid they just ease everyone in on the second question of the entire SATs paper. Still, I’ll let them off, because a) yes, children should be able to carry during an addition - although presenting it as an uninterrupted sea of 5s feels like an open invitation for a careless mistake; and b) they made up for it in the rest of the test.

Arithmetic paper 2026: Question 3

They knew exactly what they were doing by putting this in as the third question, didn’t they? Did anyone else get looks from their kids when they reached this one?

Arithmetic paper 2026: Question 25

Anyone who knows me knows how much I love teaching long division (sad, I know, but are you surprised at this point?), and it is such a satisfying method when the dividend and divisor are quite complex numbers. However, as sad as I was not to have to use the long division method, this question was chef’s kiss because it could be solved mentally with a bit of number sense. Of course, that relies on children actually analysing the calculation and choosing an efficient method, rather than spotting a two-digit divisor and immediately launching into listing multiples.

Arithmetic 2026: Question 26

Here’s another example of the test assessing whether children automatically reach for a memorised method or whether they can think mathematically about the concept itself. Dividing a fraction by an integer is often taught by multiplying the reciprocal; so here, children may have rewritten the calculation as ¾ × ⅓. However, those with a deeper understanding would recognise that splitting three quarters into three equal parts simply gives one quarter.

The first method still leads to the correct answer (even if some children may not simplify the resulting fraction), but the children thinking more flexibly about the maths would likely reach it far more efficiently.

Arithmetic 2026: Question 30

Once again, this was a brilliant question for distinguishing confident mathematicians from those relying purely on procedure. One of the most efficient approaches would be to find 1% of 400 and subtract three lots of it from 400 - something that can be done entirely mentally. However, I wouldn’t be surprised if many children instead built up 97% by combining 9 lots of 10% with 7 lots of 1%. There’s absolutely nothing wrong with that method, but it is far more time-consuming, potentially costing children valuable time later in the paper. (Personally, I think spotting that 97% of 400 is simply 97 × 4 is just as satisfying.)

Arithmetic 2026: Question 31

Flexible thinking or method-dependency? This could absolutely be solved using the classic long division algorithm, but factor pairs offer a much more elegant route. Two quick divisions (either by 8 then 2, or by 4 twice) would reach the same answer, and probably much faster too.

Arithmetic 2026: Question 36

Oof… I just know this question will have caught some children out by leaving their answer as a fraction instead of simplifying it to a whole number. A commonly taught (and entirely valid) method is to convert the mixed number into an improper fraction before multiplying, so this becomes 5/2 × 70/1, giving 350/2. However, the mark scheme expects answers equivalent to integers to be simplified fully - in this case, 175.

Children with stronger number sense, though, may have ‘seen’ the calculation differently: two lots of 70 is 140, and half of 70 is 35, giving 175 far more efficiently than the formal fraction method.

Overall, this year’s arithmetic paper felt far less about executing formal written methods and far more about recognising when not to use them. Children with strong number sense and flexible thinking would likely have found several opportunities to save valuable time, whereas those relying purely on rehearsed procedures may have struggled to work efficiently under pressure. In many ways, the paper rewarded mathematical confidence just as much as mathematical accuracy.

Interestingly, that theme didn’t disappear in Papers 2 and 3 either.

Want the 2026 GPS insights?

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

Read more

Content domain analysis of Papers 2 and 3: reasoning

In the 2025 maths paper analysis, we noted that Scale Factors (R3) and Estimate, Measure and Read Scales (M2) hadn’t appeared for four years now, so could be worth prioritising for revision. Another year on, they still have not featured, meaning questions based on these substrands have now been absent since before the two-year gap in SATs papers during Covid.

Other substrands that may be worth prioritising in future are Missing Number Problems Expressed in Algebra (A1) and Circles (G5), neither of which has appeared in the last three papers. Despite Calculations maintaining a similar overall weighting to previous years, only six of the nine available substrands were assessed across the three papers: Add/Subtract Mentally (C1), Estimate, Use Inverses and Check (C3), and Order of Operations (C9) were all absent.

It was also the first time in several years that Equivalent Fractions (F2) did not appear, and the first time ever that Convert Between Metric Units (M5) was absent from the paper. While some questions may initially seem to fit these content domains, checking against the official mark scheme confirms that this was not the case.

Analysis of maths Paper 2

Wasn’t this first reasoning paper a charmer! They clearly lulled us into a false sense of security with that arithmetic test.

Teacher response to maths Paper 2

A small minority of teachers thought this paper was ‘easy’, whilst the rest of you, well…

How did teachers find the 2026 maths reasoning paper 2? Nearly 1000 teachers voted…

Did anyone else have hideous weather during SATs week? Totally unfair considering one of the few perks of this week is the opportunity to at least get outside for a game of rounders!

The majority of you thought your children found it difficult - and once we dive into the questions, it’s clear why.

How did teachers think their pupils found the 2026 maths reasoning paper 2? Nearly 1,000 teachers voted…

Teachers also said:

• ‘So many steps for 1 mark. Lots ran out of time’
• ‘I excelled in GCSE and A-level maths and thought it was tough on 11 year olds’
• ‘Naughty how they seemed to not graduate the level of Qs’
• ‘I had children in tears. Qs were tough and aimed more at GDS. Horrid!’
• ‘So many tears - extra timers even giving up because they just felt defeated by it’
• ‘Overall thought the phrasing was harsh and not fair to children with decent maths but poor reading!’

Let’s take a look at some of the more noteworthy questions.

Reasoning 2026: Paper 2, Question 7

When learning about position and direction, children are taught the vocabulary of ‘translation’ and ‘reflection’. They should also learn the umbrella term ‘transformation’ (which encompasses both), but it would be very easy for some pupils to confuse ‘transformation’ with ‘translation’.

In this question, four transformations of Shape A are shown, three of which are reflections (one in the x-axis, one in the y-axis and one in the line x = -4.5). The correct answer is therefore the shape in the top-right quadrant, as it is the only image that has been translated rather than reflected. However, I can easily imagine some children being thrown by the wording here.

That being said, I really liked this question - it was a lovely little test of transformation vocabulary.

Reasoning 2026: Paper 2, Question 12

The common thread continues: distinguishing between those who can round based on a memorised rule against those who have a good conceptual understanding. I suspect some pupils may have been thrown by the fact that two of the rounded answers were identical - something that feels ‘wrong’ if you’re relying purely on procedure rather than thinking carefully about the value of the number.

Reasoning 2026: Paper 2, Question 16

Fractions - everyone’s favourite! At least the first part was relatively kind, with denominators that were straightforward multiples of the originals. However, asking children to convert five separate fractions into fortieths felt quite laborious for a single mark. Some pupils may have spotted more efficient comparisons conceptually without fully converting - for example, recognising that 4/8 is equivalent to ½, which is greater than ⅖.

Reasoning 2026: Paper 2, Question 20

This was one of the most controversial questions in the entire paper. Many of you seemed to dislike it (or at least despair over the ways your pupils approached it), but - don’t hate me - I actually thought it was a brilliant question. To reach the correct answer, children needed to work systematically, think logically and stay accurate throughout: essentially all my favourite things about maths.

Some teachers raised concerns that certain children may barely have encountered 5p or 2p coins in real life (which is arguably a separate issue in itself), but for me, this question was far less about money and far more about the underlying algebraic thinking.

Reasoning 2026: Paper 2, Question 21

Now this one I didn’t like. This is clearly testing their knowledge of ratio, but they could at least have chosen a slightly friendlier number than 3.5 for the calculations. Dividing 3.5 litres by 2 gives 1.75 litres (and I suspect some children may even have converted to millilitres first for perceived ‘ease’, turning it into 3500 ÷ 2 = 1750ml), before then needing to multiply by 3 to reach 5.25 litres. Any children working in millilitres would also have needed to remember to convert back to litres at the end. All that for just 2 marks!

Reasoning 2026: Paper 2, Question 23

The Marmite question! I was part of the 35% I’m afraid.

How did teachers find the last question on the 2026 maths reasoning paper 2?

However, I do think it should have been three marks. The children had to:

• convert fractions into percentages (⅕ = 20%, ¼ = 25%)
• correctly combine those percentages (20% + 25% + 30% + 20% = 95%)
• recognise from the pie chart that the remaining sector, swimming, must therefore represent 5%
• then use ratio and proportional reasoning to determine the whole amount, knowing that 5% represented three children

You’re not telling me that shouldn’t deserve three marks.

Analysis of the 2026 maths Paper 3

What a pleasant surprise Paper 3 was. It does, however, make me question the logic behind the order of the reasoning papers. Would it have been better to ease children in with the more accessible paper first, or is it nicer for them to finish the run of tests feeling more confident after a comparatively gentler paper? 

Teacher response to maths paper 3

How did teachers think the 2026 maths reasoning paper 3 compared to reasoning paper 2?

Some mixed feedback, but overall, teachers seemed to prefer it to Paper 2:

• ‘I thought it was the best of all reasoning papers ever… until I looked over their shoulders🤣’
• ‘Questions were friendlier but still a lot to do’
• ‘Less wordy & more accessible for my lower attainers. Finished feeling success :)’
• ‘No tears today!’
• ‘A fairly nice paper. Nice to see some less regular topics’
• ‘Compared to yesterday, it was a breeze!’

Reasoning 2026, Paper 3 Question 13

This would have been a lovely candidate for a bar model (although I’m an algebra girl myself). Thankfully, the mark allocation here actually reflected the amount of thinking required: one mark for recognising that Box A represented a quarter of 600g (150g), and another for calculating that Box B is three times that amount (450g). 

Reasoning 2026: Paper 3, Question 14

This one took me a minute! And I love it when a question makes me stop and think. I divided 126 by 7 (18), then found factor pairs of the answer (18 and 1, 6 and 3, or 9 and 2). Would you have done it differently?

Reasoning 2026: Paper 3, Question 19

This question was well-placed near the end as it’s a little trickier than many of the others; however, a great opportunity to show some mathematical thinking - and not unlike the 2p/5p question from Paper 2!

Reasoning 2026: Paper 3, Question 21

We finally get our three-marker! And I like it - clear wording, clear steps, no ‘tricky’ numbers. As long as children knew how to calculate the area of a rectangle and a triangle (although, for technical accuracy, I would have preferred the right angle to have been marked) and were reasonably fluent with the four operations, they were onto a winner.

KS2 SATs maths papers: results

The ‘pass mark’ has historically ranged from 49% (2024) to 55% (2018), 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 in maths) has remained within a much narrower range of 85% to 87%. 

Based on previous year’s data, I’d expect the ‘pass’ percentage to be in the low 50s again… but time will tell!

Maths 2026: conclusion

Regardless of opinions on the papers themselves, SATs week is always an enormous achievement for year 6 children and teachers alike. So, whether your class floated out of Paper 3 feeling triumphant or crawled over the finish line powered purely by SATs-breakfast-croissants, you all deserve a huge congratulations.