Top grades come from the hardest marks: multi-step problems, extended evaluation and unfamiliar contexts. Schools raise A* and grade 9 rates by giving high attainers regular practice on those question types, feedback against the top of the mark scheme, and the right tier entry. Setting alone does little: the EEF rates it at about one month’s gain for high attainers.
How do A* and grade 9 work?
Through thresholds set each series, not fixed percentages. Cambridge says “a grade threshold is the minimum number of marks that a candidate needs to achieve to be awarded a particular grade”, and “the thresholds may change from one series to the next” (Cambridge). At Pearson Edexcel International GCSE, “the new grade 9 represents a new level of attainment and has been introduced to differentiate your top performing students”, and “the bottom of the grade 7 broadly aligns with the bottom of the grade A” (Pearson). A 9 is therefore harder to reach than an A.
How many students get A* worldwide?
It varies widely by subject. From Cambridge’s June 2026 world statistics:
| Subject | % A* worldwide, June 2026 |
|---|---|
| IGCSE Additional Mathematics | 26.1 |
| A Level Further Mathematics | 24.2 |
| IGCSE Physics | 20.9 |
| IGCSE Mathematics | 18.9 |
| A Level Mathematics | 16.9 |
| A Level Chemistry | 16.5 |
| IGCSE First Language English | 14.6 |
| A Level Economics | 12.9 |
Sources: Cambridge A Level and IGCSE results statistics, June 2026. Compare your department’s A* rate with the world figure for the same syllabus, not a school-wide target. See Cambridge June 2026 results.
What does the evidence say about grouping high attainers?
That setting helps them a little and costs low attainers more. The EEF Toolkit rates setting and streaming at “0 months” overall, “no impact for very low cost based on very limited evidence”. For pupils with high prior attainment it gives +1: impact is “similar to, or slightly higher than, the impact on all pupils overall”. Low attainers in sets “make about 3 months less progress”. Other approaches rate higher:
| EEF Toolkit approach | Impact | Evidence rating |
|---|---|---|
| Setting and streaming | 0 months (+1 for high attainers) | “Very limited evidence” |
| Individualised instruction | +4 months | ”Limited evidence”; secondary studies “tend to show higher effects” |
| Mastery learning | +5 months | ”Limited evidence”; students must show mastery, “typically 80%”, before moving on |
| Feedback | +6 months | ”Extensive evidence” |
| Metacognition and self-regulation | +8 months | ”Extensive evidence” |
So what happens inside the class matters more than which class a student is in.
Which strategies stretch high attainers?
Seven we recommend, drawing on that evidence:
- Target the top-mark questions. Identify which questions separate A from A* in recent papers: the last parts of structured questions, and the extended-response and evaluation items. Practise those weekly.
- Feed back against the top band. Show high attainers what the full-mark answer contains and where theirs fell short, using mark schemes and examiner reports.
- Teach metacognition explicitly. Have students plan, check and correct their own answers against the mark scheme, with the EEF’s +8 months in mind.
- Move on only after mastery. High attainers still need secure foundations; check with a short test before accelerating.
- Use individualised practice. Let high attainers work on harder material at their own pace, which the EEF finds works better with secondary-age pupils.
- Get the entry right. Cambridge Extended, or Higher tier in Pearson Mathematics A, so the top grade is available. See Core vs Extended.
- Add depth, not only speed. Additional Mathematics, Further Mathematics, the Project Qualification or Global Perspectives stretch without rushing.
How should a department find its A* candidates?
From mock data, not reputation. Use a full mock on the series’ own boundaries, then list students within a few marks of the A* or grade 9 threshold. Those students, not only the highest scorers, gain most from targeted practice. See how to set grade boundaries for mocks and question-level analysis.
What should leaders avoid?
- Acceleration without depth, such as early entry before a student is ready for the top grade.
- Setting as the only strategy, given the EEF’s limited evidence of gains.
- School-wide A targets* that ignore big differences between subjects.
- Neglecting low attainers in the process; the EEF’s -3 months for low attainers in sets is a real cost.
How schools do this with AI Buddy
High attainers need a steady supply of harder questions and fast feedback. AI Buddy gives them past-paper-style practice with instant marking, with adaptive practice that moves students on as they show mastery. Teachers see which students are close to the A* or grade 9 threshold by topic, and which question types hold them back.
Frequently asked questions
How do schools increase A* grades at IGCSE?
By practising the question types that separate A from A*, giving feedback against the top of the mark scheme, teaching students to check their own work, and entering students for the tier where A* is available.
What percentage of students get A* at IGCSE?
It depends on the subject. In June 2026, 18.9% of Cambridge IGCSE Mathematics candidates worldwide achieved A*, and 20.9% in Physics.
Is grade 9 the same as A*?
No. Pearson says grade 9 “represents a new level of attainment” to differentiate top students; the bottom of grade 7 broadly aligns with the bottom of grade A.
Does setting help high attainers?
Slightly. The EEF rates the effect for high attainers at about +1 month, with very limited evidence, while low attainers in sets make about 3 months less progress.
Discover how AI Buddy helps schools strengthen teaching, learning and evidence-informed school improvement. Or start a short consultation with our schools team using the form below — we will get back to you directly.
Sources
- Cambridge International, Grade threshold tables and results statistics for A Level and IGCSE, June 2026
- Pearson, International GCSE 9–1 grading scale
- Education Endowment Foundation, Teaching and Learning Toolkit: Setting and streaming, Mastery learning, Individualised instruction, Feedback and Metacognition and self-regulation