Master the IB Computer Science HL exam on the 2027 guide. Both paper structures, command terms, marking criteria, trace-table technique, and how to write the code that earns the marks — including the abstract data types SL never studies.
240 teaching hours • 2 external papers • 1 internal assessment (computational solution)
The same two papers as SL, sat for two hours each and carrying 40% apiece — which is why the internal assessment counts for 20% at HL rather than 30%.
Two sections sat together, 80 marks. Section A (56 marks) is extended response on theme A at SL and HL — the SL questions reappear with extra parts assessing HL-only material. Section B (24 marks) asks questions on the pre-seen case study, which at HL carries four challenge questions to research rather than the two at SL.
What to expect:
Key Tips
Easy Marks
Watch Out
Extended, scenario-based questions on theme B at SL and HL, 80 marks. Two questions are repeated from the SL paper, including one on algorithmic thinking with no code to read or write, and the rest focus on HL-only topics such as recursion and abstract data types. The paper comes in a Python version and a Java version; sit the one your school teaches.
What to expect:
Key Tips
Easy Marks
Watch Out
Command terms tell you exactly what the examiner expects. Filter by Assessment Objective (AO).
Match your answer depth to the marks available.
Example questions:
One precise line. Padding earns nothing and wastes time you need later.
Example questions:
Count the marks and make that many distinct points — one point per mark.
Example questions:
Justify against the alternative — saying what would go wrong with the option you rejected is usually where the last marks are.
These concepts appear throughout Computer Science HL exams. Master them to score higher.
Outline wants the shape of the idea; explain wants the mechanism and its consequence; evaluate wants a judgement. Writing more than the term asks for earns nothing extra.
Write the place values or the division remainders. A wrong final answer with correct method still earns method marks; a bare wrong number earns none.
One column per variable, one row per step, plus a column for the loop condition. It is slow on purpose — the value is that you never guess what happens next.
Empty input, a single item, and the value exactly on the boundary. Off-by-one errors are the commonest lost marks in Paper 2.
Learn from others' mistakes. These cost students marks every exam session.
Treating "explain" as "outline" — naming a thing without its mechanism.
Say what happens AND what follows from it. Two sentences, not one.
Comparing two things by describing each separately.
Write point against point: "A does X, whereas B does Y."
Losing an element at the start or end of a loop.
Check the bound on an example: an array of n has indices 0 to n − 1.
Dividing to find an average without checking for an empty list.
Handle the empty case first — it is the edge case that breaks every average.
Using AND where the logic needs OR when testing an invalid range.
A valid range needs AND; testing for invalid needs OR. "< 0 AND > 100" is never true.
Normalising a table without naming the dependency that was removed.
State it explicitly: "ProductName depends on ProductID, part of the key — a partial dependency."
20% of final grade • Developed over 35 hours
A working solution built for a real client, documented from the problem definition through design and development to testing and evaluation against your own success criteria.
Marking Criteria
Tips for Top Marks
Apply these exam skills with our Computer Science HL practice questions. Get instant AI feedback that shows exactly what scored marks and how to improve.